FINGERPRINT BASED ATM SYSTEM

by

Md. Mashud Rana

Roll: CSE 05106476
Jannatul Ferdausi
Roll: CSE 05106482

A Project Submitted in Partial Fulfillment of the Requirements for the Degree of

Bachelor of Science in Computer Science & Engineering

DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING

STAMFORD UNIVERSITY BANGLADESH

November 2016
 DECLARATION

We, hereby, declare that the work presented in this Project is the outcome of the
investigation performed by us under the supervision of Mr. Tamjid Rahman, Senior
Lecturer, Department of Computer Science & Engineering, Stamford University
Bangladesh. We also declare that no part of this Project and thereof has been or is being
submitted elsewhere for the award of any degree or Diploma.

Countersigned Signature

…………………………….. ….………………………

(Mr. Tamjid Rahman) (Md. Mashud Rana)

Supervisor …………………………

(Jannatul Ferdausi)

Candidates
 ABSTRACT

Biometrics technology is rapidly progressing and offers attractive opportunities. In recent

years, biometric authentication has grown in popularity as a means of personal identification
in ATM authentication systems. Fingerprint Based ATM is a desktop application where
fingerprint of the user is used as an authentication. The finger print minutiae features are
different for each human being so the user can be identified uniquely. Instead of using ATM
card, Fingerprint based ATM is safer and secure. There is no worry of losing ATM card and
no need to carry ATM card in one‘s wallet. Fingerprint is just need to be used in order to do
any banking transaction. The user has to login using his/her fingerprint and he/she has to
enter the pin code in order to do further transaction. The user can withdraw money from his
account. User can transfer money to various accounts by mentioning account number. In
order to withdraw money user has to enter the amount he wants to withdraw and has to
mention from which account he wants to withdraw (i.e. saving account, current account) .The
user must have appropriate balance in his ATM account to do transaction. User can view the
balance available in his respective account. The system will provide the user to view last 5
transactions and others transaction can also be seen by permission of Bank.

iii
 ACKNOWLEDGEMENTS
First of all we would like to thank the almighty ALLAH. Today we are successful in
completing our work with such ease because He gave us the ability, chance, and cooperating
supervisor.

We would like to take the opportunity to express our gratitude to Mr. Tamjid Rahman, our
respected supervisor. Although he was always loaded with several other activities, he gave us
more than enough time in this work. He not only gave us time but also proper guidance and
valuable advice whenever we faced with some difficulties. His comments and guidance
helped us in preparing our project report.

We would like to thank Mr Tamjid Rahman, our respected teacher, who inspired us in every
step. We are also thankful to our teachers who helped us in a number of ways by providing
various resources and moral support.

Last of all we are grateful to our family who are, always with us in our every step of life.

iv
TABLE OF CONTENTS
ABSTRACT iii

ACKNOWLEDGMENTS iv
TABLE OF CONTENTS v
LIST OF FIGURES ix
LIST OF TABLES xi

1
Chapter 1: Introduction
1.1 Introduction……………………………………………………….......... 2
1.2 Literature Review……………………………………………………….. 3
1.3 Outline of this project…...……………………………………………… 4

Chapter 2: ATM Transaction 5

2.1 What is ATM?..................………………………………………………. 6

2.2 Internal Structure of ATM…………………………………………….. 6
2.3 Feature of ATM Machines………………….…………………………. 7
2.4 ATM Transaction Process……………..………………………………. 8
2.5 ATM Cell Structure…………….…………………………………......... 10
2.6 ATM Protocol……………….……………………………………......... 12
2.7 Summary of this chapter……...……………………………………….. 13

14

Chapter 3: ATM Connection

3.1 Connection Type of ATM………….………………………………….. 15
3.2 Connection process of ATM…………………………………………... 15
3.2.1 ATM Network Design……………………………………………. 15
3.2.2 ATM Signaling…………….. ……………………………………... 16
3.2.3 Point-to-Point Connection……………………………...………… 16
3.2.4 Point-to-Point Multipoint Connection……………………………. 17
3.3 Vendor of ATM………………………………………………………... 18
3.3.1 Hardware Manufacture of ATM…………………………………... 15
18
3.3.2 Software Manufacturer of ATM…………………………………... 15
18
3.3.3 Ways of Transaction……………………………………………… 18

v
 3.4 Network………………………………………………………….…….. 19
3.4.1 Network provider……………………………………….………… 19
3.4.2 Bank’s Under This Network…………………………..…………… 19
3.4.3 Connectivity Type………………………………….……………… 19
3.4.4 Cable type……………………………………….………………… 19
3.5 Transaction Process……………………………….……………………. 20
3.5.1 Steps of Transaction Process……………..………………………... 20
3.5.2 Network Maintain……………………..…………………………... 20
3.5.3 Connectivity Type…………………..……………………………... 21
3.6 Cable Type…………………………..………………………………….. 15
21
3.7 Summary of this chapter………...……..……………………………….. 15
22

Chapter 4: Biometric Techniques 23

4.1 Biometric Techniques………………………………………………... 24

4.1.1 Fingerprint Technologies….……………………………………. 24
4.2 Fingerprint Readers…………….…………………………...………... 24
4.3 Fingerprint processing………………….……………………………. 26
4.4 Hand Geometry………………..…………………………………….. 28
4.5 Other Biometric Technologies………………………………………. 29
4.6 Biometrics and Cryptography..………………………………………. 31
4.7 Summary of this chapter..……………………………………………. 31

Chapter 5: Identification 32

5.1 Improving security With Biometrics.………………………………… 33

5.1.1 Need For Automation…………..………………………………. 33
5.1.2 Early AFIS Development………..……………………………… 34
5.1.3 FBI AFIS Initiative………………..…………………………….. 34
5.1.4 French AFIS Initiative…………….…………………………..... 36
5.1.5 United Kingdom AFIS Initiative….…………………………….. 36

5.1.6 Japanese AFIS Initiative………………………………………….. 37

5.1.7 The Politicization of Fingerprints and the San Francisco……………. 43
38
5.1.8 AFIS Proliferation……………………………………………...… 3
39
5.2 AFIS Operation…………………………………………………….... 40
5.2.1 AFIS Function and Capabilities……………………………….…. 40
..Experiment
5.2.2 Technical Function……………………………………………… 41

5.2.3 System Accuracy…………………………………………………. 42

35
vi
 5.2.3.1 Community Safety……....…… …….…………………..... 43
5.2.3.2 Validation of Friction Ridge Science………………….….. 3744

5.2.4 IAFIS………………….…………………………………………. 45
5.2.4.1 IAFIS Status as of Early 2006……………………………... 43
48
5.2.4.2 Universal Latent Work Station……….…………………...... 349
5.3 Summary of this chapter……………………...………………….…... 45
49
37

Chapter 6: Digitization and Processing of Fingerprint 50

6.1 Digitization and Processing of Fingerprints…………………………. 51

6.1.1 Algorithms…………..…………………………………………. 51
6.1.2 Image Acquisition……………………………………………… 52
6.1.3 Image Enhancement…………………………………………….. 54
6.1.4 Enhancement off Latent Prints for AFIS Searching……………. 55
6.1.5 Automated Enhancement of Fingerprint Image………………. 56

6.2 Feature Extraction…….…………………………………………… 57

6.3 Matching…………………………………………………………….. 43
60
6.3.1 Indexing and Retrieval…………………………………………. 3
63
6.4 Fingerprint Standards…………………………………………… 66
6.4.1 Record Types………………...…………………………………. 67
..Experiment
6.4.2 Finger Image Quality…………………………….……………… 68
6.5 Latent Interoperability……………………………………………… 69
6.6 Summary of this chapter…………………….……………………… 70

Chapter 7: Biometric Security Using Finger print Recognition 71

7.1 Biometric Security Using Finger print Recognition…………….. 72

7.1.1 Problems in collecting Fingerprints……………………………... 73
7.1.2 Ink-based Fingerprint……………..……………………………... 75
7.1.3 Optical Methods………………………………………………… 76
7.1.4 Thermal Imaging………………………………….……………. 78
7.1.5 Electromagnetic field imaging………………………………….... 79

7.1.6 Ultrasound imaging……………………………………………….. 80

7.1.7 Comparison testing…………….……………………………………… 43
82
7.2 Summary of this chapter………………………………………………. 83

3
..Experiment
vii
 84

Chapter 8: Proposed system Fingerprint For ATM

8.1 Fingerprint For ATM System………………………………….. 85
8.2 Fingerprint Authenticated ATM System……………………………... 86
8.3 For a customer bifurcation point………………….…………………. 87
8.4 The SEPLA Protocol………………………………….……………. 90
8.5 Biometric Enrollment……………..…………………………………. 94

8.6 Finger Scan technology for ATM………………………………………… 95

8.7 Power Supply for ATM………………………………………………….. 43
97
8.8 Summary of this chapter…………………………………………….. 397

..Experiment
Appendix 99
References 3128

..Experiment

viii
 LIST OF FIGURE

2.1 Shows the Complete Internal Structure……………………………………… 6

2.2 ATM Machines……………………….……………………………………… 7

2.3 Interactive Components of ATM…………………………………………….. 8

2.4 ATM Transaction Process…….……………………………………………… 9

2.5 X.25 ATM Network Design………………………………………………….. 13

3.1 Basic Connection of ATM…………………………………………………… 15

3.2 ATM Signaling……………………………….……………………………… 16

3.3 Transaction in ATM Network………………………………………………... 20

3.4 Block Diagram of ATM Transaction Process………………………………... 21

4.1 Optical fingerprint Reader…………………………………….……………... 25

4.2 Fingerprint Bitmap…………………………..……………………………….. 25

4.3 Ultra-Scan Fingerprint Bitmap……………………………..………………... 26

4.4 PRIP MSU Process……………………………………….………………….. 27

4.5 2D Hand Shape………………………………………………………………. 28

4.6 Hand Scan Recognition System……………………………………………... 29

5.1 Tracking latent hits through the court……………..………………………… 39

5.2 Statistical study of AFIS hits vs. burglaries in San Francisco…………………. 40

5.3 Illustration from the Federal Bureau of Investigation………………………. 46

5.4 Integrated Automated Finger Identification System….. …………………….. 47

5.5 IAFIS Network Architecture…………………………………………………. 47

6.1 Automatic fingerprint-capture algorithm ……………………………………. 51

6.2 Live Scan…………………………………………………………………….. 53

Finger Image Enhancement ………………………………………………..... 55

6.3

ix
6.4 Enhancement of Latent Finger Prints …..…………………………………… 56

6.5 Contextual filtering-based Finger Image Enhancement Algorithm………….. 58

6.6 Stages in a typical fingerprint minutiae extraction algorithm………………... 59

6.7 Stages in a typical fingerprint minutiae matching algorithm………………… 61

6.8 Six commonly used fingerprint classes ………..………………………….. 65

7.1 Problems in collecting fingerprints ……………………..…………………… 74

7.2 Optical method of Fingerprint……………………………………………… 77

7.3 Thermal Imaging…………..………………….……………………………… 78

7.4 Ultrasound Imaging...……….………………………………………………... 81

7.5 Image Comparison………………………….………………………………... 82

8.1 Fingerprint for ATM System……………. …………………………………... 85

8.2 Fingerprint authenticated ATM system…………….............................................. 86

8.3 For a customer bifurcation point……………………………………………... 88

8.4 The SELPA Protocol……….. ……………………………..………………... 92

8.5 Biometric Enrollment………………………………………………………… 94

8.6 Finger scan technology for ATM…………………………………………….. 96

8.7 Power supply for ATM………..…………………………………………….. 97

x
 LIST OF TABLES

5.1 Minimum hits from 10 largest states by population for 2005 42

xi
 CHAPTER 1
INTRODUCTION
1.1 Introduction
Biometrics are automated methods of recognizing a person based on a physiological or
behavioral characteristic. Biometric-based solutions are able to provide for confidential
financial transactions and personal data privacy. The various features used are face,
fingerprints, hand geometry, handwriting, iris, retina, vein and voice. Fingerprinting or
finger-scanning technologies are the oldest of the biometric sciences and utilize distinctive
features of the fingerprint to identify or verify the identity of individuals. Finger-scan
technology is the most commonly deployed biometric technology, used in a broad range of
physical access and logical access applications. All fingerprints have unique characteristics
and patterns. A normal fingerprint pattern is made up of lines and spaces. These lines are
called ridges while the spaces between the ridges are called valleys. It is through the pattern
of these ridges and valleys that a unique fingerprint is matched for verification and
authorization. These unique fingerprint traits are termed ―minutiae‖ and comparisons are
made based on these traits. On average, a typical live scan produces 40 ―minutiae‖. The
Federal Bureau of Investigation (FBI) has reported that no more than 8 common minutiae can
be shared by two individuals. On most modern cash machines, the customer is identified by
inserting a plastic ATM card with a magnetic stripe or a plastic smart card with a chip that
contains a unique card number and some security information such as an expiration date
or CVVC (CVV). Authentication is provided by the customer entering a personal
identification number (PIN).

Using a cash machine, customers can access their bank deposit or credit accounts in order to
make a variety of transactions such as cash withdrawals, check balances, or credit mobile
phones. If the currency being withdrawn from the cash machine is different from that in
which the bank account is denominated the money will be converted at an official exchange
rate. Thus, cash machines often provide the best possible exchange rates for foreign
travellers, and are widely used for this purpose. Prior to the industrial revolution and the
mass migrations to the cities, populations lived mostly in rural communities where everyone
knew everyone else and there was little need for identification. Indeed, there were no police
forces, no penitentiaries, and very few courts. As cities became crowded, crime rates soared
and criminals flourished within a sea of anonymity. Newspapers feasted on stories of
lawlessness, legislatures quickly responded with more laws and harsher penalties (especially

2
for repeat offenders), and police departments were charged with identifying and arresting the
miscreants. Identification systems—rogues‘ galleries, anthropometry, Bertillon‘s ―portrait
Parle‖, and the Henry system—emerged and quickly spread worldwide at the end of the 19th
and beginning of the 20th century.

The late 1960s and early 1970s witnessed another era of civil turmoil and an unprecedented
rise in crime rates, but this era happened to coincide with the development of the silicon chip.
The challenges inherent in identification systems seemed ready-made for the solutions of
automatic data processing, and AFIS—Automated Fingerprint Identification System—was
born.

1.2 Literature Review

Biometric systems have overtime served as robust security mechanisms in various domains.
Fingerprints are the newest and most widely used form of biometric identification. The use of
fingerprint for identification has been employed in law enforcement for about a century. A
much broader application of fingerprint is for personal authentication, for instance to access a
computer, a network, an ATM machine, a car or a home.
The method uses 3D analysis of the finger for tracking and identification purposes. An
individual places their hand (palm down) onto a special plate. [3] A camera takes a picture of
it and analyzes the length, width, thickness and surface area of the hand. This recorded
biostatistics information is then stored for future use. Companies have used this type of
biometrics for attendance tracking and accessing secure entrances.

All the biometrics, fingerprint recognition is one of the most reliable and promising personal
identification technologies. Fingerprints are plays an important role in biometric system. In
biometrics technologies, fingerprint authentication has been in use for the longest time and
bears more advantages than other biometric technologies do. Fingerprints are the most
widely used biometric feature for person identification and verification. But in this paper we
proposed that fingerprint verification of ATM (Automatic Teller Machine) security system
using the biometric with hybridization. The fingerprint trait is chosen, because of its
availability, reliability and high accuracy. The fingerprint based biometric system can be
implemented easily for secure the ATM machine. In this system the working of these ATM
machine is when the customer place on the fingerprint module when it access the ATM for
draw the cash then, the machine wants to fingerprint of that user‘s which use the machine.

3
Using biometric, it verify/identify fingerprint and gives accurate result that if it valid or not
valid. In this way we can try to control the crime circle of ATM and do secure it.

The growth in electronic transactions has resulted in a greater demand for fast and accurate
user identification and authentication. Access codes for buildings, banks accounts and
computer systems often use personal identification numbers (PIN's) for identification and
security clearances. Conventional method of identification based on possession of ID cards
or exclusive knowledge like a social security number or a password are not all together
reliable. An embedded fingerprint biometric authentication scheme for automated teller
machine (ATM) banking systems is proposed in this paper. In this scheme, a fingerprint
biometric technique is fused with the ATM for person authentication to ameliorate the
security level.

1.3 Outline of the Project

Identification and verification of a person today is a common thing; which may include door-
lock system, safe box and vehicle control or even at accessing bank accounts via ATM, etc.
which is necessary for securing personal information. The conventional methods like ID card
verification or signature does not provide perfection and reliability. The systems employed at
these places must be fast enough and robust too. Use of the ATM (Automatic Teller
Machine) which provides customers with the convenient banknote trading is facing a new
challenge to carry on the valid identity to the customer. Since, in conventional identification
methods with ATM, criminal cases are increasing making financial losses to customers. A
simple fingerprint recognition system using LPC2148 as a core controller. The system uses
fingerprint scanner to capture fingerprints with its DSP processor and optical sensor. This
system can be employed at any application with enhanced security because of the uniqueness
of fingerprints. It is convenient due to its low power requirement and portability.

4
 CHAPTER 2
ATM TRANSACTION
2.1 What is ATM?
ATM is Automated Teller Machine. [1] Now it‘s making peoples life very easy as they get
their money when they need. So, they do not need to carry either big amount of money or the
cheek book all the time. To get rid from this burden they need to deposit money in the bank
by opening an account and then the bank will be given a Card i.e. an ATM card with a PIN
number to them. By using that they can withdraw money from any ATM machine of that
bank. When they insert the card in the machine and the PIN number the machine will show
few instructions on the screen. By that time verification (PIN Number and Account Number)
will be done with the main bank computer as they are connected. If the verification is correct
then the user will choose an instruction and the ATM will dispense money to the card holder.

2.2 Internal Structure of ATM

In the following pictures we have the internal structure of two different type of ATM
machine. And also it can be divided into two different parts:
- Upper Unit
- Lower Unit

Figure-2.1: Shows the Complete Internal Structure.

6
In the upper unit it has the CPU that processes and validates customer details by connecting
to the bank computer after customer has entered ATM card. This ATM has few layered
boxes in the lower unit of it. These boxes are called currency boxes or cassettes where
currencies are kept for withdrawal or the deposited money to be kept. A rubber roller is there
to check if more than one banknote is moving and also sensor to see that more than one
banknote or bill stuck together or not when cash is dispensing. There is a receipt printer as
we see in the figure to print current statistics of the cardholder‘s account or every times
cardholder withdraws cash.

2.3 Figure of ATM Machines

Figure-2.2: ATM machines

Card Reader: Customer inserts their card in it when there is written ―Please Insert your
card‖ on the screen.
Keypad: Use for PIN code input, choices, amount of money etc. as the input to the ATM
machine.
Display Screen: This screen shows all the instructions or options for the customers‘
convenience.

7
Screen Buttons: When options are given on the screen one user can choose any of the
options accordingly by the use of button on left or right side of the screen. These buttons
select the option from the screen.
Cash Dispenser: Withdrawal money is given by this slot.
Deposit Slot: To deposit money this slot is use.
Speaker: Speaker provides the facilities to the customer by giving auditory feedback.

Figure-2.3: Interactive components of ATM

2.4 ATM Transaction Process

The main objective of this system is to develop a system, which is for ATM security
applications. [9] In this system Bankers will collect the customers finger prints and mobile
number while opening the accounts then customer only access ATM machine. The working
of these ATM machine is when customer place finger on the finger print module when it
access automatically generates every time different 4-digit code as a message to the mobile
of the authorized customer through GSM modem connected to the microcontroller. The code

8
receive by the customer should be entered by pressing the keys by the customer should be
entered by pressing the keys on the screen. After entering it checks whether it is a valid one
or not and allows the customer further access. ATMs are mainly used for transaction such as
cash withdrawal, money transfer and payment of telephone bills and electricity bills. The
working of the existing system is depicted in the below figure.

Figure-2.4: ATM Transaction Process

Personal Identification Number (PIN) provides security in Existing ATM system. PIN is a
four digit number and is generated by the respective financial institution. A user can change
his/her PIN. But, that day as the code tracking is increased, the PIN strength is decreased.
The existing system the user has to insert the card and the PIN number in the ATM System.
The System allows for transaction only if the PIN is correct Else, the system asks for PIN
again and a maximum of three times is allowed. Now-a-days, in the self-service banking
system has got extensive popularization with characteristic offering high-quality 24 hours
service for customer. Using the ATM which provide customer with the financial crime case
rises repeatedly in recent years, a lot of criminals tamper with the ATM terminal and steal
user‘s credit card and password by illegal means. Once user‘s bank card is lost and the
password is stolen, the criminal will draw all cash in the shortest time, which will bring
enormous financial losses to customer.

9
2.5 ATM Cell Structure

At either a private or a public user-network interface (UNI), an ATM cell always consists of
a 5-byte header followed by a 48-byte payload. The header is composed of six elements, each
detailed in Figure

Generic Flow Control

The Generic Flow Control (GFC) field is a 4-bit field that was originally added to support the
connection of ATM networks to shared access networks such as a Distributed Queue Dual
Bus (DQDB) ring. The GFC field was designed to give the User-Network Interface (UNI) 4
bits in which to negotiate multiplexing and flow control among the cells of various ATM
connections. However, the use and exact values of the GFC field have not been standardized,
and the field is always set to 0000.

Virtual Path Identifier

The Virtual Path Identifier (VPI) defines the virtual path for this particular cell. VPIs for a
particular virtual channel connection are discovered during the connection setup process for
switched virtual circuit (SVC) connections and manually configured for permanent virtual
circuit (PVC) connections. At the UNI, the VPI length of 8 bits allows up to 256 different
virtual paths. VPI 0 exists by default on all ATM equipment and is used for administrative
purposes such as signaling to create and delete dynamic ATM connections.

Virtual Channel Identifier

The Virtual Channel Identifier (VCI) defines the virtual channel within the specified virtual
path for this particular cell. Just as with VPIs, VCIs are also discovered during the
connection setup process for switched virtual circuit (SVC) connections and manually
configured for permanent virtual circuit (PVC) connections. The VCI length of 16 bits allows
up to 65,536 different virtual channels for each virtual path. VCIs 0 to 15 are reserved by the
ITU and VCIs from16 to 32 are reserved by the ATM Forum (for each virtual path). These
reserved VCIs are used for signaling, operation and maintenance, and resource management.
The combination of VPI and VCI values identifies the virtual circuit for a specified ATM
cell. The VPI/VCI combination provides the ATM forwarding information that the ATM

10
switch uses to forward the cell to its destination. The VPI/VCI combination is not a network
‗layer address such as an IP or IPX network address.

The VPI/VCI combination acts as a local identifier of a virtual circuit and is similar to the
Logical Channel Number in X.25 and the Data Link Connection Identifier (DLCI) in Frame
Relay. At any particular ATM endpoint or switch, the VPI/VCI uniquely identifies a virtual
circuit to the next ATM endpoint or switch. The VPI/VCI pair need not match the VCI/VPI
used by the final destination ATM endpoint.

The VPI/VCI combination is unique for each transmission path (that is, for each cable or
connection to the ATM switch). However, two different virtual circuits on two different ports
on an ATM switch can have the same VPI/VCI without conflict.

Payload Type Indicator

The Payload Type Indicator (PTI) is a 3-bit field. Its bits are used as follows:
The first bit indicates the type of ATM cell that follows. A first bit set to 0 indicates user
data; a bit set to 1 indicates operations, administration & management (OA&M) data.

The second bit indicates whether the cell experienced congestion in its journey from source
to destination. This bit is also called the Explicit Forward Congestion Indication (EFCI) bit.
The second bit is set to 0 by the source; if an interim switch experiences congestion while
routing the cell, it sets the bit to 1. [10] After it is set to 1, all other switches in the path leave
this bit value at 1.

Destination ATM endpoints can use the EFCI bit to implement flow control mechanisms to
throttle back on the transmission rate until cells with an EFCI bit set to 0 are received.
The third bit indicates the last cell in a block for AAL5 in user ATM cells. For non-user
ATM cells, the third bit is used for OA&M functions.

Cell Loss Priority

The Cell Loss Priority (CLP) field is a 1-bit field used as a priority indicator. When it is set to
0, the cell is high priority and interim switches must make every effort to forward the cell

11
successfully. When the CLP bit is set to 1, the interim switches sometimes discard the cell in
congestion situations. The CLP bit is very similar to the Discard Eligibility (DE) bit in Frame
Relay.

An ATM endpoint sets the CLP bit to 1 when a cell is created to indicate a lower priority
cell. The ATM switch can set the CLP to 1 if the cell exceeds the negotiated parameters of
the virtual channel connection. This is similar to bursting above the Committed Information
Rate (CIR) in Frame Relay.

Header Error Check

The Header Error Check (HEC) field is an 8-bit field that allows an ATM switch or ATM
endpoint to correct a single-bit error or to detect multi-bit errors in the first 4 bytes of the
ATM header. Multi-bit error cells are silently discarded. The HEC only checks the ATM
header and not the ATM payload. Checking the payload for errors is the responsibility of
upper layer protocols.

2.6 ATM Protocol

Protocol X.25
The protocol use for ATM connection is X.25. It is a packet switch data network protocol
which defines an international recommendation for the exchange of data as well as control
information between two end systems.
X.25 network devices fall into three general categories:
• Data terminal equipment (DTE).
• Data circuit equipment (DCE).
• Packet switching exchange (PSE).
• DTE devices : PC or network hosts (subscribers)
• DCE devices : modem, packet switches
• PSN : are switches & transfer data to DTE to DTE

12
 Figure-2.5: X.25 Network Design

Working Layer of X.25

X.25 works on lower three layers of OSI (open system interconnection) defined by ISO.
– Packet level protocol
• Similar to data link layer of OSI model
– Link level (Link access procedure balanced)
• Similar to physical layer of OSI model
– Physical level
• Similar to physical layer of OSI model

2.7 Summary of this chapter

ATM transaction processing is a core function of the ATM business equation. This Topic
Center focuses on ATM transaction processing and the companies who provide processing
services.

13
 CHAPTER 3
ATM CONNECTION
3.1 Connection Type of ATM
Different types of ATM Connections:
ATM connections mainly have two types:
- Dial up Connection using Modem
- Leased Line Connection
Dial up connection is mostly used because dial up connection is less costly. But the
throughput rate is low as it is not connected all the time. [5,7] And Leased Line Connection
is mostly use by those where throughput rate high is strongly needed. But it is costly.

3.2. Connection Process of ATM

3.2.1 ATM Network Design
ATM is connected to host computer and the host computer is connected to the Bank
Computer. Here the connection network is telephone network that may be leased line or dial
up using modem. Some places where Output is very important but the cost is not a factor
there leased line is used and the dial up connection is used where cost is important. Host
computer mainly work as a gateway between ATM and bank computer. Many ATM‘s can be
connected through this host computer.

Figure-3.1: Basic Connection of ATM

15
Basically ATM machines are connected to host computer and host computer is connected to
Bank computer. Basically host computer is the third party which maintains all these facilities
and equipment. Sometimes ATM machines are directly connected to Bank computer.

3.2.2 Signaling
Signaling components exist at the end station and at the ATM switch. The signaling layer of
ATM software is responsible for creating, managing, and terminating switched virtual
circuits (SVCs). The ATM standard wire protocol implemented by the signaling software is
called the User Network Interface (UNI). The way one ATM switch signals another ATM
switch comprises a second signaling standard, called the Network Interface (NNI).

Figure 3.2 : ATM Signaling

3.2.3 Point-to-Point Connection

When an ATM-aware process seeks to connect to another process elsewhere on the network,
it asks the signaling software to establish an SVC. To do this, the signaling software sends an
SVC creation request to the ATM switch using the ATM adapter and the reserved signaling
VC. Each ATM switch forwards the request to another switch until the request reaches its
destination. An ATM switch determines which switch to send the request to next based on
the ATM address for the connection and the switch's internal network database (routing
tables). Each switch also determines whether or not the request's service category and

16
Quality of Service needs can be met. At any point in this process, a switch can refuse the
request.

If all the switches along the path can support the virtual circuit as requested, the destination
end station receives a packet that contains the VC number. From that point on, the ATM-
aware process can communicate with the destination process directly by sending packets to
the VPI/VCI that identify the specified VC.

The ATM adapter shapes data traffic for each VC to match the contract made with the ATM
network. If too much data is sent for any reason, the ATM switch can ignore — and lose —
the data in favor of providing bandwidth to another contract or set of contracts. This is true
for the entire breadth of the network; if bandwidth or speed exceeds the limits established by
the contract, any device, including the ATM adapter, can simply drop the data. If this
happens, the end stations concerned are not notified of the cell loss.

3.2.4 Point-to-Multipoint Connection

Unlike a standard LAN environment, ATM is a connection-oriented medium that has no

inherent capabilities for broadcasting or multicasting packets. To provide this ability, the
sending node can create a virtual circuit to all destinations and send a copy of the data on
each virtual circuit. However, this is highly inefficient. A more efficient way to do this is
through point-to-multipoint connections. [9] Point-to-multipoint connects a single source
endpoint, known as the root node, to multiple destination endpoints, known as leaves.
Wherever the connection splits into two or more branches, the ATM switches copy cells to
the multiple destinations.
Point-to-multipoint connections are unidirectional; the root can transmit to the leaves, but the
leaves cannot transmit to the root or to each other on the same connection. Leaf-to-node and
leaf-to-leaf transmission requires a separate connection. One reason for this limitation is the
simplicity of AAL5 and the inability to interleave cells from multiple payloads on a single
connection.

17
3.3. Vendor of ATM

3.3.1 Hardware Manufacturer of ATM

The following manufacturers mainly supply the complete ATM‘s.
- NCR (National cash Register)
- IBM (International Business Machines)
- Diebold was an American financial self-service, security and services corporation
- Tidel Has one of the finest integrated infrastructure for the Information Technology
industry, matching global standards.
- Triton Manufacturer and supplier of ATM ventilation and window products

3.3.2 Software Manufacturer of ATM

Software manufacturer of ATM machines is mainly
- KAL is a global software company dedicated solely to multivendor ATM software. KAL's
product suite enables ATM hardware, software and services sourced from multiple vendors
to work together perfectly. This allows banks to adopt a best-of-breed strategy, reduce costs,
increase functionality and provide an enhanced customer experience.

3.3.3 Ways of Transaction

In Bank Asia transactions are divided into three main categories
◊ My bank to others bank
A customer of a bank uses other banks ATM.
◊ Others bank to my bank
Other banks customer uses ATM of Bank Asia.
◊ My bank to my bank
A customer uses it‘s own bank ATM machine.

18
3.4 Network

3.4.1 Network Provider

ETN is the network provider to all these nine banks. Total ATM maintenance, equipment
facilities are provided by ETN. Also card issue, report are provided by ETN.

3.4.2 Bank’s Under This Network

This network contains nine banks.
National Bank Ltd.
NCC Bank Ltd.
Dhaka Bank Ltd.
Social Investment Bank Ltd.
Islami Bank Ltd.
Bank Asia Ltd.
South East Bank Ltd.
Agroni Bank Ltd.
Commercial Bank of Ltd.

3.4.3 Connectivity Type

For ATM machines Bank Asia (provided by ETN) uses two types of connections. They use
leased line for connection. Bank Asia uses two connection lines because if one line is down
immediately other one will be activated within a minute. These connections are:
◊ DDN (under T&T)
◊ Metro net

3.4.4 Cable Type

Bank Asia (provided by ETN) uses fiber optic cable for ATM connection. As Fiber optic
given highest data rate that is why they use this type of connection.

19
3.5 Transaction Process

Figure-3.3: Transaction in Network

3.5.1. Steps of Transaction Process

First customer insert ATM card (E-cash card) into the machine and wait to insert PIN
(personal identification number). When both processes are done ATM Machine check
account number and PIN for further processing like requesting money to the bank server.
Bank Server debited the amount of money from the customer account. [10] And update
database for that customer account and send all transaction information to ETN server. ETN
server then update database so that they can send report to the banks. And then ETN send
clearance signal to the ATM machine to dispenser. After the clearance signal ATM machine
dispense money to the customer.

3.5.2. Network Maintain

BRAC Bank provides all their ATM network solutions by themselves. They have their own
strong IT department and their duty is to look after the entire ATM network through out the
whole nation. They do not use any third party company for there ATM support.

20
3.5.3 Connectivity Type
BRAC Bank basically uses the Lease Line connectivity for their entire network system.
They also have the backup connectivity option so that if one line goes down then another one
will be automatically up within a minute. [12] This gives them tremendous support for their
reliable banking services.

3.6. Cable Type

Bank mainly uses Fiber Optic as Fiber Optic provides highest data rates.

Figure-3.4: Block Diagram of ATM Transaction Process

Step 1: Customer Insert ATM card and PIN number. Then ATM machines verify account
number and PIN number for further processing. If verification is successful, ATM machine
takes dispense amount and forward those information and dispense amount to the Bank
Server.

Step 2: Again Bank Server verifies the amount and if successful sends signal to the ATM
machine.

Step 3: If the machine gets successful signal from the Bank Server then it dispenses money
to the customer. Otherwise on not successful signal machine shows error message.

21
3.7 Summary of this chapter
The ATM adapter shapes data traffic for each VC to match the contract made with the ATM
network. If too much data is sent for any reason, the ATM switch can ignore — and lose —
the data in favor of providing bandwidth to another contract or set of contracts. This is true
for the entire breadth of the network; if bandwidth or speed exceeds the limits established by
the contract, any device, including the ATM adapter, can simply drop the data. If this
happens, the end stations concerned are not notified of the cell loss.

22
 CHAPTER 4
BIOMETRIC TECHNIQUES
4.1 Biometric Techniques
There are lots of biometric techniques available nowadays. A few of them are in the stage of
the research only (e.g. the odor analysis), but a significant number of technologies is already
mature and commercially available (at least ten different types of biometrics are
commercially available nowadays: fingerprint, finger geometry, hand geometry, palm print,
iris pattern, retina pattern, facial recognition, voice comparison, signature dynamics and
typing rhythm).

4.1.1 Fingerprint technologies

Fingerprint identification is perhaps the oldest of all the biometric techniques. Fingerprints
were used already in the Old China as a means of positively identifying a person as an author
of the document. Their use in law enforcement since the last century is well known and
actually let to an association fingerprint crime. This caused some worries about the user
acceptance of fingerprint-based systems. The situation improves as these systems spread
around and become more common. Systems that can automatically check details of a
person‘s fingerprint have been in use since the 1960s by law enforcement agencies. The U.S.
Government commissioned a study by Sandia Labs to compare various biometric
technologies used for identification in early seventies. This study concluded that the
fingerprint technologies had the greatest potential to produce the best identification accuracy.
The study is quit outdated now, but it turned the research and development focus on the
fingerprint technology since its release.

4.2 Fingerprint readers

Before we can precede any further we need to obtain the digitalized fingerprint. The
traditional method uses the ink to get the fingerprint onto a piece of paper. This piece of
paper is then scanned using a traditional scanner. This method is used only rarely today when
an old paper-based database is being digitalized, a fingerprint found on a scene of a crime is
being processed or in law enforcement AFIS systems. Otherwise modern live fingerprint
readers are used. They do not require the ink anymore. These live fingerprint readers are
most commonly based on optical, thermal, silicon or ultrasonic principles. Optical fingerprint
readers are the most common at present. They are based on reflection changes at the spots

24
where the finger papillary lines touch the reader‘s surface. The size of the optical fingerprint
readers typically is around 10X10X5 centimeters. It is difficult to minimize them much more
as the reader has to comprise the source of light, reflection surface and the light sensor.

‗
Figure-4.1: Optical fingerprint Reader
The optical fingerprint readers work usually reliably, but sometimes have problems with dust
if heavily used and not cleaned. The dust may cause latent fingerprints, which may be
accepted by the reader as a real fingerprint. Optical fingerprint readers cannot be fooled by a
simple picture of a fingerprint, but any 3D fingerprint model makes a significant problem, all
the reader checks is the pressure. A few readers are therefore equipped with additional
detectors of finger licenses.

Figure- 4.2 : Fingerprint Bitmap

Optical readers are relatively cheap and are manufactured by a great number of
manufacturers. The field of optical technologies attracts many newly established firms (e.g.,
American Biometric Company, Digital Persona) as well as a few big and well-known
companies (such as HP, Philips or Sony). Optical fingerprint readers are also often embedded
in keyboards, mice or monitors. Both optical and silicon fingerprint readers are fast enough
tol capture and display the fingerprint in real time. The typical resolution is around 500 DPI.

25
 Figure – 4.3: Ultra-Scan Fingerprint Bitmap
Ultrasonic fingerprint readers are the newest and least common. They use ultrasound to
monitor the finger surface. The user places the finger on a piece of glass and the ultrasonic
sensor moves and reads whole the fingerprint. This process takes one or two seconds.
Ultrasound is not disturbed by the dirt on the fingers so the quality of the bitmap obtained is
usually fair. Ultrasonic fingerprint readers are manufactured by a single company nowadays.
This company (Ultra Scan Inc.) owns multiple patents for the ultrasonic technology. The
readers produced by this company are relatively big (15X15X20 centimeters), heavy, noisy
and expensive (with the price around $2500). They are able to scan fingerprints at 300, 600
and 1000 DPI (according to the model).

4.3 Fingerprint processing

Fingerprints are not compared and usually also not stored as bitmaps. Fingerprint matching
techniques can be placed into two categories: minutiae-based and correlation based.
Minutiae-based techniques find the minutiae points first and then map their relative
placement on the finger. Minutiae are individual unique characteristics within the fingerprint
pattern such as ridge endings, bifurcations, divergences, dots or islands (see the picture on
the following page). [16] In the recent years automated fingerprint comparisons have been
most often based on minutiae. The problem with minutiae is that it is difficult to extract the
minutiae points accurately when the fingerprint is of low quality. This method also does not
take into account the global pattern of ridges and furrows. The correlation-based method is
able to overcome some of the difficulties of the minutiae-based approach. However, it has

26
some of its own shortcomings. Correlation-based techniques require the precise location of a
registration point and are affected by image translation and rotation. The readability of a
fingerprint depends on a variety of work and environmental factors. These include age,
gender, occupation and race. A young, female, Asian mineworker is seen as the most difficult
subject. A surprisingly high proportion of the population have missing fingers, with the left
forefinger having the highest percentage at 0.62%.

Figure -4.4: PRIP MSU Process

There are about 30 minutiae within a typical fingerprint image obtained by a live fingerprint
reader. The FBI has shown that no two individuals can have more than 8 common minutiae.
The U.S. Court system has allowed testimony based on 12 matching minutiae. The number
and spatial distribution of minutiae varies according to the quality of the fingerprint image,
finger pressure, moisture and placement. In the decision process, the biometric system tries to
find a minutiae transformation between the current distribution and the stored template. The
matching decision is then based on the possibility and complexity of the necessary
transformation. The decision usually takes from 5 milliseconds to 2 seconds. The speed of
the decision sometimes depends on the security level and the negative answer very often
takes longer time than the positive one (sometimes even 10 times more). There is no direct
dependency between the speed and accuracy of the matching algorithm according to our
experience. We have seen fast and accurate as well as slow and less accurate matching
algorithms. The minutiae found in the fingerprint image are also used to store the fingerprint
for future comparisons. The minutiae are encoded and often also compressed. The size of
such a master template usually is between 24 bytes and one kilobyte. Fingerprints contain a
large amount of data. Because of the high level of data present in the image, it is possible to
eliminate false matches and reduce the number of possible matches to a small fraction. This
means that the fingerprint technology can be used for identification even within large
databases. Fingerprint identification technology has undergone an extensive research and

27
development since the seventies. The initial reason for the effort was the response to the FBI
requirement for an identification search system. Such systems are called Automated
Fingerprint Identification Systems (AFIS) and are used to identify individuals in large
databases (typically to find the offender of a crime according to a fingerprint found at the
crime scene or to identify a person whose identity is unknown). [ 8] AFIS systems are
operated by professionals who manually intervene the minutiae extraction and matching
process and thus their results are really excellent. In today‘s criminal justice applications, the
AFIS systems achieve over 98% identification rate while the FAR is below 1%. The typical
access control systems, on the other side, are completely automated. Their accuracy is
slightly worse. The quality of the fingerprint image obtained by an automated fingerprint
reader from an inexperienced (non-professional) user is usually lower.

4.4 Hand Geometry

Hand geometry is based on the fact that nearly every person‘s hand is shaped differently and
that the shape of a person‘s hand does not change after certain age. Hand geometry systems
produce estimates of certain measurements of the hand such as the length and the width of

Figure 4.5 : 2D Hand Shape

Various methods are used to measure the hand. These methods are most commonly based
either on mechanical or optical principle. The latter ones are much more common today.
Optical hand geometry scanners capture the image of the hand and using the image edge
detection algorithm compute the hand‘s characteristics. There are basically 2 subcategories of
optical scanners. Devices from the first category create a black and white bitmap image of
the hand‘s shape. This easily has done using a source of light and a black-and white camera.
The bitmap image is then processed by the computer software. Only 2D characteristics of the

28
hand can be used in this case. Hand geometry systems from the other category are more
sophisticated. They use special guide markings to position the hand better and have two.

Figure 4.6 : Hand Scan Recognition System

Hand geometry scanners are easy to use. Where the hand must be placed accurately, guide
markings have been incorporated and the units are mounted so that they are at a comfortable
height for majority of the population. The noise factors such as dirt and grease do not pose a
serious problem, as only the silhouette of the hand shape is important. [9,12] The only
problem with hand geometry scanners is in the countries where the public do not like to place
their hand down flat on a surface where someone else‘s hand has been placed. A few hand
geometry scanners produce only the video signal with the hand shape. Image digitalization
and processing is then done in the computer. On the other side there exist very sophisticated
and automated scanners that do everything by themselves including the enrollment, data
storage, verification and even simple networking with a master device and multiple slave
scanners. The size of a typical hand geometry scanner is considerably big (30 ∨ 30 ∨ 50 cm).
This is usually not a problem as the hand geometry scanners are typically used for physical
access control (e.g. at a door), where the size is not a crucial parameter.

4.5 Other Biometric Technologies

Palm print: Palm print verification is a slightly different implementation of the fingerprint
technology. [15] Palm print scanning uses optical readers that are very similar to those used
for fingerprint scanning, their size is, however, much bigger and this is a limiting factor for
the use in workstations or mobile devices.

29
Hand vein: Hand vein geometry is based on the fact that the vein pattern is distinctive for
various individuals. The veins under the skin absorb infrared light and thus have a darker
pattern on the image of the hand taken by an infrared camera. The hand vein geometry is still
in the stage of research and development. One such system is manufactured by British
Technology Group. The device is called Vein check and uses a template with the size of 50
bytes.

DNA: DNA sampling is rather intrusive at present and requires a form of tissue, blood or
other bodily sample. This method of capture still has to be refined. So far the DNA analysis
has not been sufficiently automatic to rank the DNA analysis as a biometric technology. The
analysis of human DNA is now possible within 10 minutes. As soon as the technology
advances so that DNA can be matched automatically in real time, it may become more
significant. At present DNA is very entrenched in crime detection and so will remain in the
law enforcement area for the time being.

Thermal imaging: This technology is similar to the hand vein geometry. It also uses an
infrared source of light and camera to produce an image of the vein pattern in the face or in
the wrist.

Ear shape: Identifying individuals by the ear shape is used in law enforcement applications
where ear markings are found at crime scenes. Whether this technology will progress to
access control applications is yet to be seen. An ear shape verifier (Otophone) is produced by
a French company ART Techniques. It is a telephone-type handset within which is a lighting
unit and cameras which capture two images of the ear.

Body odor: The body odor biometrics is based on the fact that virtually each human smell is
unique. [12] The smell is captured by sensors that are capable to obtain the odor from non-
intrusive parts of the body such as the back of the hand. Methods of capturing a person‘s
smell are being explored by Mastiff Electronic Systems. Each human smell is made up of
chemicals known as volatiles. They are extracted by the system and converted into a
template. The use of body odor sensors brings up the privacy issue as the body odor carries a

30
significal amount of sensitive personal information. It is possible to diagnose some diseases
or activities in the last hours (like sex, for example) by analyzing the body odor.
Keystroke dynamics: Keystroke dynamics is a method of verifying the identity of an
individual by their typing rhythm which can cope with trained typists as well as the amateur
two-finger typist. Systems can verify the user at the log-on stage or they can continually
monitor the typist. These systems should be cheap to install as all that is needed is a software
package.
Fingernail bed: The US company AIMS is developing a system which scans the dermal
structure under the fingernail. [11] This tongue and groove structure is made up of nearly
parallel rows of vascular rich skin. Between these parallel dermal structures are narrow
channels, and it is the distance between these which is measured by the AIMS system.

4.6 Biometrics and Cryptography

Is cryptography necessary for the secure use of biometric systems? The answer is quite clear:
Yes. There are basically two kinds of biometric systems:

*Automated identification systems operated by professionals. The purpose of such systems is

to identify an individual in question or to find an offender of a crime according to trails left
on the crime scene. The operators of these systems do not have any reason to cheat the
system, so the only task for the cryptography is to secure the sensitive biometric data.

*Access control systems. These systems are used by ordinary users to gain a privilege or an
access right. Securing such a system is much more complicated task. Let us consider further
the general-use systems of the latter type, as this report is devoted solely to the use of
biometrics for the authentication.

4.7 Summary of this chapter

Biometrics refers to metrics related to human characteristics. Biometrics authentication (or
realistic authentication) is used in computer science as a form of identification and access
control. It is also used to identify individuals in groups that are under surveillance. People
Who Read This Also Read. Fingerprint recognition in the fight against football hooliganism.
Biometric authentication achoice for banks

31
 CHAPTER 5

IDENTIFICATION
5.1 Improving security with biometrics
Can biometrics help cryptography to increase the security? Here the answer is not so clear.
Cryptography has been relatively successfully used without biometrics over decades. But it
still can benefit from the use of biometrics. To put it simple, cryptography is based on keys.
Secure storage of keys is a crucial non-trivial task. Key management often is the weakest
point of many systems. Secret and private keys must be kept secret, and here the biometric
technologies might help. Indeed, one of the most promising applications of biometrics is the
secret key protection. If a user‘s local workstation is trusted, then the problem of the
authentication software is minor, but the input device must be trustworthy. The security
concerns are the same no matter whether the secret (or private) keys are stored on a
smartcard or on the hard disk of the workstation. If a user‘s workstation is not trusted, the
private keys have to be stored in a separate secure place, usually a smartcard. Smartcard
based solutions where the secret key is unlocked only after a successful biometric
verification increase the overall security, as the biometric data does not need to leave the
card. For smartcards the fingerprint techniques with a silicon fingerprint reader are most
commonly used today.

5.1.1 Need For Automation

In 1924, the FBI‘s Identification Division was established by authority of the United States
congressional budget appropriation bill for the Department of Justice. The identification
division was created to provide a central repository of criminal identification data for law
enforcement agencies throughout the United States. The original collection of fingerprint
records contained 810,188 records. [14] After its creation, hundreds of thousands of new
records were added to this collection yearly, and by the early 1960s the FBI‘s criminal file
had grown to about 15 million individuals. This was in addition to the 63 million records in
the civilian file, much of which was the result of military additions from World War II and
the Korean conflict.

Almost all of the criminal file‘s 15 million individuals contained 10 rolled fingerprints per
card for a total of 150 million single fingerprints. Incoming records were manually classified
and searched against this file using the FBI‘s modified Henry system of classification.
Approximately 30,000 cards were searched daily. The time and human resources to
accomplish this daily workload continued to grow. As a card entered the system, a

33
preliminary gross pattern classification was assigned to each fingerprint by technicians. The
technicians could complete approximately 100 fingerprint cards per hour. Complete
classification and searching against the massive files could only be accomplished at an
average rate of 3.3 cards per employee per hour. Obviously, as the size of the criminal file
and the daily workload increased, the amount of resources required continued to grow.
Eventually, classification extensions were added to reduce the portion of the criminal file that
needed to be searched against each card. Nonetheless, the manual system used for searching
and matching fingerprints was approaching the point of being unable to handle the daily
workload.

5.1.2 Early AFIS Development

In the early 1960s, the FBI in the United States, the Home Office in the United Kingdom,
Paris Police in France, and the Japanese National Police initiated projects to develop
automated fingerprint identification systems. The thrust of this research was to use emerging
electronic digital computers to assist or replace the labor-intensive processes of classifying,
searching, and matching ten print cards used for personal identification.

5.1.3 FBI AFIS Initiative

By 1963, Special Agent Carl Volker of the FBI‘s Identification Division realized that the
manual searching of the criminal file would not remain feasible for much longer. In an
attempt to resolve this problem, he sought the help of engineers Raymond Moore and Joe
Western of the National Institute of Standards and Technology (NIST). After describing his
problem, he asked for assistance in automating the FBI‘s fingerprint identification process.

The NIST engineers first studied the manual methods used by human fingerprint technicians
to make identifications. These methods were based on comparing the minutiae (i.e., ridge
endings and ridge bifurcations) on fingerprint ridges. If the minutiae from two fingerprints
were determined to be topologically equivalent, the two fingerprints were declared to be
identical—that is, having been recorded from the same finger of the same person. After this
review, and after studying additional problems inherent with the inking process, they
believed that a computerized solution to automatically match and pair minutiae could be
developed that would operate in a manner similar to the techniques used by human examiners

34
to make fingerprint identifications. But to achieve this goal, three major tasks would have to
be accomplished. First, a scanner had to be developed that could automatically read and
electronically capture the inked fingerprint image. Second, it was necessary to accurately and
consistently detect and identify minutiae existing in the captured image. Finally, a method
had to be developed to compare two lists of minutiae descriptors to determine whether they
both most likely came from the same finger of the same individual.

The Identification Division of the FBI decided that the approach suggested by Moore and
western should be followed. To address the first two of the three tasks, on December 16,
1966, the FBI issued a Request for Quotation (RFQ) ―for developing, demonstrating, and
testing a device for reading certain fingerprint minutiae‖ (FBI, 1966). This contract was for a
device to automatically locate and determine the relative position and orientation of the
specified minutiae in individual fingerprints on standard fingerprint cards to be used for
testing by the FBI. The requirements stated that the reader must be able to measure and
locate minutiae in units of not more than 0.1 mm and that the direction of each minutiae must
be measured and presented as output in units of not more than 11.25 degrees (1/32 of a full
circle). The initial requirements called for a prototype model to process 10,000 single
fingerprints (1,000 cards). Contractors were also instructed to develop a proposal for a
subsequent contract to process 10 times that number of fingerprints.

The 14 proposals received in response to this RFQ were divided into 5 broad technical
approaches. At the conclusion of the proposal evaluation, two separate proposals were
funded to provide a basic model for reading fingerprint images and extracting minutiae. Both
proposed to use a ―flying spot scanner‖ for capturing the image. But each offered a different
approach for processing the captured image data, and both seemed promising. One contract
was awarded to Cornell Aeronautical Labs, Inc., which proposed using a general-purpose
digital computer to process binary pixels and develop programs for detecting and providing
measurement parameters for each identified minutiae. The second contract was awarded to
North American Aviation, Inc., Automatic‘s Division, which proposed using a special-
purpose digital process to compare fixed logical marks to the image for identifying,
detecting, and encoding each minutia.

35
5.1.4 French AFIS Initiative
In 1969, M. R. Thiebault, Prefecture of Police in Paris, reported on the French efforts.
(Descriptions of work done by Thiebault can be found in the entries listed in the Additional
Information section of this chapter.) France‘s focus was on the solution to the latent
fingerprint problem rather than the general identification problem that was the concern in the
United States. The French approach incorporated a vision (a video camera tube) to scan
photographic film transparencies of fingerprints. Scanning was done at 400 pixels per inch
(ppi), which was less than an optimal scan rate for latent work. This minutiae matching
approach was based on special-purpose, high-speed hardware that used an array of logical
circuits. [5] The French also were interested in resolving the problem of poor fingerprint
image quality. In order to acquire a high-contrast image that would be easy to photograph
and process, a technique was developed to record live fingerprint images photographically
using a principle of ―frustrated total internal reflection‖ (FTIR). Although not put into large-
scale production at that time, 20 years later FTIR became the cornerstone for the
development of the modern-day live scan fingerprint scanners. These are making the use of
ink and cards obsolete for non-forensic identification purposes today.

By the early 1970s, the personnel responsible for development of France‘s fingerprint
automation technology had changed. As a result, there was little interest in pursuing
automated fingerprint identification research for the next several years. In the late 1970s, a
computer engineering subsidiary of France‘s largest financial institution responded to a
request by the French Ministry of Interior to work on automated fingerprint processing for
the French National Police. [7] Later, this company joined with the Morphologic
Mathematics Laboratory at the Paris School of Mines to form a subsidiary called Morpho
Systems that went on to develop a functioning. Currently, Morpho Systems is part of Sagem
(also known as Group SAFRAN).

5.1.5 United Kingdom AFIS Initiative

During the same period of time, the United Kingdom‘s Home Office was doing research into
automatic fingerprint identification. Two of the main individuals responsible for the United
Kingdom‘s AFIS were Dr. Barry Blain and Ken Millard. (Papers produced by Millard are

36
listed in the Additional Information section of this chapter). Like the French, their main focus
was latent print work. By 1974, research was being done in-house with contractor assistance
from Ferranti, Ltd. [2] The Home Office developed a reader to detect minutiae, record
position and orientation, and determine ridge counts to the five nearest neighbors to the right
of each minutia. This was the first use of ridge count information by an AFIS vendor.

5.1.6 Japanese AFIS Initiative

Like France and the United Kingdom, Japan‘s motivation for a fingerprint identification
system was directed toward the matching of latent images against a master file of rolled
fingerprints. Japan‘s researchers believed that an accurate latent system would naturally lead
to the development of an accurate ten print system.

Within a few years, the fingerprint automation focus of Japanese researchers had changed.
By 1969, the Identification Section of the Criminal Investigation Bureau, National Police
Agency of Japan (NPA), approached NEC to develop a system for the computerization of
fingerprint identification. NEC determined that it could build an automated fingerprint
identification system employing a similar minutiae-based approach to that being used in the
FBI system under development. At that time, it was thought that a fully automated system for
searching fingerprints would not be realized for 5 to 10 years. In 1969, NEC and NPA
representatives visited the FBI and began to learn about the current state of the art for the
FBI‘s AFIS plans. During the same period, NPA representatives also collaborated with
Moore and Wegstein from NIST. Additional AFIS sites were visited where information was
acquired regarding useful and worthless approaches that had been attempted. All of this
information was evaluated and used in the development of the NEC system.

For the next 10 years, NEC worked to develop its AFIS. In addition to minutiae location and
orientation, this system also incorporated ridge-count information present in the local four
surrounding quadrants of each minutiae under consideration for pairing. By 1982, NEC had
successfully installed its system in the NPA and started the card conversion process. Within a
year, latent inquiry searches began.

37
5.1.7 The Politicization of Fingerprints and the San
Francisco Experiment
Early development and implementation of automated fingerprint systems was limited to
national police agencies in Europe, North America, and Japan. But the problems associated
with huge national databases and the newborn status of computer technology in the 1970s
limited the utility of these systems. Government investment in AFIS was justified largely on
the promise of efficiency in the processing of incoming tenprint records. But funding these
expensive systems on the local level would demand some creativity.

Following the success of the FBI‘s Finder, Rockwell took its system to market in the mid-
1970s. Ken Moses of the San Francisco Police Department had attended several of those
Printpack conferences and became a staunch crusader for fingerprint automation. In three
successive years, he persuaded the Chief of Police to include a Printpack system in the city
budget, but each time it was vetoed by the mayor. After the third mayoral veto, a ballot
proposition was organized by other politicians. The proposition asked citizens to vote on
whether they wanted an automated fingerprint system. [3] In 1982, Proposition E passed with
an 80% plurality.

The mayor refused to approve a sole-source purchase from Rockwell, even though it was the
only system in the world being marketed. She insisted on a competitive bid with strict
evaluation criteria and testing. While on a trade mission to Japan, the mayor learned that the
Japanese National Police were working with NEC to install a fingerprint system, but NEC
stated that the system was being developed as a public service and the company had no plans.

Besides being the first competitive bid on 1980s technology, what differentiated the San
Francisco system from those that had gone before was organizational design. AFIS was
viewed as a true system encompassing all aspects of friction ridge identification—from the
crime scene to the courtroom. The AFIS budget included laboratory and crime scene
equipment, training in all phases of forensic evidence, and even the purchase of vehicles. In
1983, a new crime scene unit was organized specifically with the new system as its
centerpiece. Significant organizational changes were put into effect:

1. All latents that met minimum criteria would be searched in AFIS.

38
 2. A new unit called Crime Scene Investigations was created and staffed on a 24/7
schedule.

3. Department policies were changed to mandate that patrol officers notify crime scene
investigators of all felonies with a potential for latent prints.

Figure 5.1 Tracking latent hits through the court(Bruton,1989)

4. All crime scene investigators who processed the crime scenes were trained in the use of
the system and en39ourage to search their own cases.

5. Performance statistics were kept from the beginning, and AFIS cases were tracked
through the criminal justice system to the courts.

At a time when burglary rates were steeply rising in cities across the nation, the burglary rate
plummeted in San Francisco (Figure 6–2; Bruton, 1989). Los Angeles even enlisted the
backing of film stars to stir up public support. The identification of serial killer Richard
Ramirez, the infamous Night Stalker, through a search of the brand-new California State
AFIS made worldwide headlines and guaranteed the future fund-ing of systems in California.

5.1.8 AFIS Proliferation

The widely publicized success in San Francisco provided the spark for the rapid proliferation
of new AFIS installations along with a methodology of benchmark testing to evaluate the
claims of the growing number of competing vendors. Governments quickly provided funding

39
so that, by 1999, the International Association for Identification‘s (IAI‘s) AFIS Directory of
Users identified 500 AFIS sites worldwide (IAI, 1999).

The burgeoning market in these multimillion-dollar systems put forensic identification on the
economic map. Commercial exhibits at IAI‘s conferences that had formerly featured
companies hawking tape and powder now expanded to digital image enhancement, lasers and
forensic light sources, and the latest in new developments from Silicon Valley. Fingermatrix
installed the first livescan device in the San Francisco Police Identification Bureau in 1988.
AFIS brought crime scene and forensic identification out of the basement; no local or state
law enforcement administrator wanted.

Figure 5.2 Statistical study of AFIS hits vs. burglaries in San Francisco.

5.2 AFIS Operations

5.2.1 AFIS Functions and Capabilities

Identification bureaus are legislatively mandated to maintain criminal history records.
Historically, this meant huge file storage requirements and cadres of clerks to maintain and
search them. Demographic-based criminal history computers were established well ahead of
AFIS, first as IBM card sort systems and then as all-digital information systems with
terminals throughout the state and, via the National Crime Information Center (NCIC)
network and the National Law Enforcement Teletype System (Nlets), throughout the nation.
These automated criminal history systems [9] became even more labor-intensive than the
paper record systems they supposedly replaced. In many systems, more paper was generated
and placed into the history jackets along with the fingerprint cards, mug shots, warrants, and
other required documents.

40
5.2.2 Technical Functions.
Law enforcement AFISs are composed of two interdependent subsystems: the tenprint (i.e.,
criminal identification) subsystem and the latent (i.e., criminal investigation) subsystem.
Each subsystem operates with a considerable amount of autonomy, and both are vital to
public safety. The tenprint subsystem is tasked with identifying sets of inked or livescan
fingerprints incident to an arrest or citation or as part of an application process to determine
whether a person has an existing record.

In many systems, identification personnel are also charged with maintaining the integrity of
the fingerprint and criminal history databases. Identification bureau staffs are generally
composed of fingerprint technicians and supporting clerical personnel. An automated tenprint
inquiry normally requires a minutiae search of only the thumbs or index fingers. Submitted
fingerprints commonly have sufficient clarity and detail to make searching of more than two
fingers unnecessary.

The search of a latent print is more tedious and time-consuming than a tenprint search. Latent
prints are often fragmentary and of poor image quality. Minutiae features are normally
reviewed one-by-one before the search begins. Depending on the portion of the database
selected to be searched and the system‘s search load, the response may take from a few
minutes to several hours to return.

Most law enforcement AFIS installations have the ability to perform the following functions:

• Search a set of known fingerprints (tenprints) against an existing tenprint database (TP–TP)
and return with results that are better than 99% accurate.

• Search a latent print from a crime scene or evidence against a tenprint database (LP–TP).

• Search a latent from a crime scene against latents on file from other crime scenes (LP–
LP).

• Search a new tenprint addition to the database against all unsolved latent prints in file
(TP–LP).

Enhancements have been developed to allow other functions that expand AFIS capabilities,
including:

41
 • Addition of palmprint records to the database to allow the search of latent palmprints
from crime scenes.

• Interfacing of AFIS with other criminal justice information systems for added efficiency
and ―lights out‖4 operation.

• Interfacing of AFIS with digital mug shot systems and livescan fingerprint capture
devices.

• Addition of hand-held portable devices for use in identity queries from the field. The
query is initiated by scanning one or more of the subject‘s fingers, extracting the minutiae
within the device, and transmitting to AFIS, which then returns a hit or no-hit (red light,
green light) result. Hit notification may be accompanied by the thumbnail image of the
subject‘s mug shot.

• Multimodal identification systems, including fingerprint, palmprint, iris, and facial

recognition, are now available.

5.2.3 System Accuracy

Most dedicated government computer systems are based on demographic data such as name,
address, date of birth, and other information derived from letters and numbers. For example,
to search for a record within the motor vehicle database, one would enter a license number or
operator data.

Table 1 Minimum hits from 10 largest states by population for 2005

1 California 8,814

2 New York 2,592

3 Illinois 1,224

4 Ohio 1,495

5 Georgia 980

Total 29,178

42
Automated fingerprint systems are based on data extracted from images. Although there is
only one correct spelling for a name in a motor vehicle database, a fingerprint image can be
scanned in an almost infinite number of ways. Success in searching fingerprints depends on
the clarity of the images and the degree of correspondence between the search print and the
database print (compression and algorithms are two other factors that can affect accuracy). In
the case of searching a new ten print card against the ten print database, there is usually more
than enough image information present to find its mate 99.9% of the time in systems with
operators on hand to check respondent lists (rather than true ―lights out‖ operations).

A latent print usually consists of a fragmentary portion of a single finger or piece of palm,
though the quality of some latent impressions can exceed their corresponding images of
record. The amount of information present in the image is usually of lesser quality and often
is contaminated with background interference. Entering latent into the computer has a
subjective element that is based on the experience of the operator. Based on latent print
acceptance test requirements commonly found in AFIS proposals and contracts, the chances
of a latent print finding its mate in the database is about 70 to 80%. Naturally, the better the
latent image, the higher the chances of success. Inversely, the chance of missing an
identification, even when the mate is in the database, is 25%. Especially in latent print
searches, failure to produce an identification or a hit does not mean the subject is not in the
database. Other factors beyond the knowledge and control of the operator, such as poor-
quality database prints, will adversely affect the chances of a match.

5.2.3.1 Community Safety.

There is no national reporting mechanism for the gathering of AFIS (or latent print) statistics, so the
measurable benefits are illusive. However, to provide some recognition of those benefits, the author
of this chapter conducted a survey of latent hits in the 10 largest states by population for the year
2005 (Table 6–1). Prior attempts to provide this type of information have revealed inconsistencies in
how identifications are counted and how the hit rate is determined (Komarinski, 2005, pp 184–189).
Based on the author‘s survey, an estimated 50,000 sus-pects a year in the United States are identified
through AFIS latent searches. In conducting the survey, if the contacted state bureaus did not have
statewide figures, attempts were made to also contact the five largest cities in that state. (In no
instance was it possible to contact every AFIS-equipped jurisdiction in a state, so the total hits are the
minimum number of hits.) Also, only case hits or suspect hits were counted, depending on what data

43
each agency kept. (When agencies reported multiple hits to a single person, this was not included in
data presented.)

Extrapolating from the table, if the remaining 40 states and all agencies of the federal
government each had just one latent hit per day, the total estimate of latent hits for the entire
United States would surpass 50,000.

Few studies have been done to measure what effect, if any, a dramatic increase in the rate of
suspect latent print identifications from AFIS has had on public safety overall. The burglary
data from San Francisco in the late 1980s (Figure 6–2) is probative but must be narrowly
construed. FBI Uniform Crime Reports show a steady decline in most serious offenses that
coincide with the proliferation of AFIS, but no cause-and-effect relationship has been
explored by academia or government. During the 1990s, many states passed ―three strikes‖
laws increasing the punishment for felony offenses that some theorists have held are respon-
sible for the decline in crime. [8] But before harsher penalties can be applied, perpetrators
must be identified and apprehended.

Burglary is the offense most impacted by AFIS. Assume that an active burglar is committing
two offenses per week when he is apprehended on the basis of an AFIS hit. He is convicted
and, based on harsh sentencing laws, sent to prison for 5 years. In this case, that one AFIS hit
will have prevented 100 crimes per year over the course of the 5 year sentence. If this one
arrest is then multiplied by some fraction of the totals from the table above, a truer apprecia-
tion of the impact that AFIS is having on society can be gained.

5.2.3.2 Validation of Friction Ridge Science.

There are many ways to test the efficacy of a theoretical proposition. Corporate and academic
laboratories pour tremendous resources into building models that they hope will closely
duplicate performance in the real world. [9] Even after successfully passing such testing,
theories fail and products get recalled after weathering the rigors of the real world. In-use
models invariably trump laboratory models.

During the past 100 years, many models have been constructed to test the theory that no two
friction ridge images from different areas of palmar surfaces are alike and to determine what
minimum number of minutiae is sufficient to support an individualization decision.

44
Automated fingerprint systems have been effectively testing identification theory millions of
times a day every day for more than 20 years. These systems tend to validate what friction
ridge examiners have propounded since Galton first set forth his standards. AFIS has also
served as a catalyst to help examiners expand their image-processing knowledge and skills.

5.2.4 IAFIS
The Integrated Automated Fingerprint Identification System, more commonly known as
IAFIS, is the world‘s largest collection of criminal history information. Fully operational
since July 28, 1999, IAFIS is maintained by the FBI‘s Criminal Justice Information Services
(CJIS) Division in Clarksburg, WV, and contains fingerprint images for more than 64 million
individuals. The FBI‘s CJIS Division system‘s architecture and the identification and
investigative services provided by the division form an integrated system-of-services (SoS)
concept. [13] These identification and information services enable local, state, federal, tribal,
and international law enforcement communities, as well as civil organizations, to efficiently
access or exchange critical information 24 hours a day, 365 days per year. The SoS provides
advanced identification and ancillary criminal justice technologies used in the identification
of subjects.

The systems within the CJIS SoS, including IAFIS, have evolved over time, both
individually and collectively, to add new technological capabilities, embrace legislative
directives, and improve the performance and accuracy of their information services. During
its first year of inception, IAFIS processed nearly 14.5 million fingerprint submissions.
Today, IAFIS processes similar tenprint volumes in as little as 3 to 4 months. Although
designed to respond to electronic criminal transactions within 2 hours and civil transactions
within 24 hours, IAFIS has exceeded these demands, often providing criminal search
requests in less than 20 minutes and civil background checks in less than 3 hours. Likewise,
IAFIS provides the latent print examiners with a superlative investigative tool, allowing
fingerprint evidence from crime scenes to be searched in approximately 2 hours rather than
the 24-hour targeted response time. Although declared a successful system early within its
deployment, IAFIS continues to improve as a vital asset to law enforcement agencies more
than 10 years later. Today‘s transient society magnifies the need for an economic, rapid,
positive identification process for both criminal and noncriminal justice background checks.
IAFIS processes are regularly improved to allow for a quick and accurate fingerprint-based

45
records check, whether related to terrorists trying to enter the United States or applicants
seeking positions of trust. Figure 6–3 illustrates the states that currently interface with IAFIS
electronically.

The increasingly complex requirements of the SoS architecture demand a well-structured

process for its operations and maintenance. Each of these systems has multiple segments
consisting of computer hardware and software that provide the operating systems and
utilities, database management, workflow management, transaction or

Figure 5.3 Illustration from the Federal Bureau of Investigation

46
 Figure 5.4 Integrated Automated Finger Identification System

messaging management, internal and external networking, communications load balancing,

and system security. IAFIS consists of three integrated segments: the Identification Tasking
and Networking (ITN) segment, the Interstate Identification Index (III), and AFIS (Figure 6–
4).

Figure 5.5 IAFIS Network Architecture

Also submitting fingerprint information to IAFIS is the Card Scanning Service (CSS). The
CSS acts as a conduit for agencies that are not yet submitting fingerprints electronically. The
CSS makes the conversion of fingerprint information from paper format to electronic format
and submits that information to IAFIS. Another system providing external communications

47
for IAFIS is Nlets. The purpose of Nlets is to provide interstate communications to law
enforcement, criminal justice, and other agencies involved in the enforcement of laws. Figure
6–5 depicts the high-level IAFIS architecture. Users wishing to interface with IAFIS
electronically must comply with the FBI‘s Electronic Fingerprint Transmission Specification
(EFTS).

5.2.4.1 IAFIS Status as of Early 2006.

Because of the evolutionary changes to the American National Standards Institute
(ANSI)/NIST standard in 1997, 2000, and 2006, the FBI has not always had the financial
resources or corporate commitment to update IAFIS and keep it current. One area where it
has moved forward is the acceptance and processing of ―segmented slaps‖ for civil
transactions. These transactions use a modified live scan platen that is 3 inches high so the
four fingers of each hand can be placed as a ―slap‖ in a straight up-and-down position.
Similarly, both thumbs can be captured simultaneously for a total of three images (type 4 or
type 14 as defined in sections 6.3.2.1 and 6.3.3). The resultant transaction‘s three-image files
are easy to segment with the capture device software. The three images and relative location
of the segmented fingers within the images are all transmitted. This dramatically reduces
collection time and improves the captured-image quality from a content perspective due to
the flat, straight, 3-inch placement.

One drawback to IAFIS is that it cannot store and search palmprints, though several
production AFISs can do so. Also, at least one foreign production and several domestic AFIS
sites accept and store 1,000-pixels-per-inch tenprint images—IAFIS cannot yet do this.

The FBI recognizes its need to expand its services and has (1) tested small palm systems and
(2) started a project known as the Next Generation Identification Program (NGI). Driven by
advances in technology, customer requirements, and growing demand for IAFIS services,
this program will further advance the FBI‘s biometric identification services, providing an
incremental replacement of current IAFIS technical capabilities while introducing new
functionality. [16] NGI improvements and new capabilities will be introduced across a
multiyear time frame within a phased approach. The NGI system will offer state-of-the-art
biometric identification services and provide a flexible framework of core capabilities that
will serve as a platform for multimodal functionality.

48
5.2.5.2 Universal Latent Work Station.
AFISs that are fully ANSI/NIST compliant can send image-based transactions from site to
site. But in the latent community, most practitioners want to edit the images and extract the
minutiae themselves, that is, perform remote searches rather than submittals. This model also
plays well with the ability of most agencies to provide the skilled labor required for imaged-
based submittals from other agencies.

The FBI CJIS Division addressed this issue by working closely with Mitretek and the four
major AFIS vendors to develop a set of tools that would permit the creation of remote
searches for any of their automated fingerprint identification systems and for IAFIS. The
result is a free software product called the Universal Latent Workstation (ULW). This
software can run on a stand-alone PC with either a flatbed scanner or a digital camera
interface. It can also run on vendor-provided latent workstations. At a minimum, when
specifying an AFIS in a procurement, one should mandate that the AFIS be able to generate
remote searches to IAFIS. It is further recommended that the procurer ask for the ability to
perform the ULW function so the vendors can integrate ULW into their systems.

5.3 Summary of this chapter

The ULW also provides the ability to launch latent print image searches into IAFIS without
the need to manually encode minutiae when working with high-quality latent prints. New to
this release is a tool that evaluates the quality of a fingerprint scan at the time it is made.
Problems such as dry skin, the size of the fingers and the quality and condition of the equip-
ment used can affect the quality of a print and its ability to be matched with other prints.

49
 CHAPTER 6
DIGITIZATION AND PROCESSING OF
FINGERPRINTS
6.1 Digitization and Processing of Fingerprints
6.1.1 Algorithms
Demands imposed by the painstaking attention needed to visually match the fingerprints of
varied qualities, the tedium of the monotonous nature of the manual work, and increasing
workloads due to a higher demand on fingerprint recognition services prompted law
enforcement agencies to initiate research into acquiring fingerprints through electronic media
and to automate fingerprint individualization based on digital representation of fingerprints.
As a result of this research, a large number of computer algorithms have been developed
during the past three decades to automatically process digital fingerprint images. [12] An
algorithm is a finite set of well-defined instructions for accomplishing some task which,
given an initial state and input, will terminate in a corresponding recognizable end-state and
output. A computer algorithm is an algorithm coded in a programming language to run on a
computer.

Figure 6.1 Automatic fingerprint-capture algorithm

51
Fingerprint images from (a) a livescan FTIR-based optical scanner; (b) a livescan capacitive scanner;
(c) a livescan piezoelectric scanner; (d) a livescan thermal scanner; (e) an off-line inked impression;
(f) a latent fingerprinthave greatly improved the operational productivity.

6.1.2 Image Acquisition

Known fingerprint data can be collected by applying a thin coating of ink over a finger and
rolling the finger from one end of the nail to the other end of the nail while pressing the
finger against a paper card. This would result in an inked ―rolled‖ fingerprint impression on
the fingerprint card. If the finger was simply pressed straight down against the paper card
instead of rolling, the resulting fingerprint impression would only contain a smaller central
area of the finger rather than the full fingerprint, resulting in an inked ―flat‖ or ―plain‖
fingerprint impression.The perspiration and contaminants on the skin result in the impression
of a finger being deposited on a surface that is touched by that finger. These ―latent‖ prints
can be chemically or physically developed and electronically captured or manually ―lifted‖
from the surface by employing certain chemical, physical, and lighting techniques. The developed
fingerprint may be lifted with tape or photographed. Often these latent fingerprints contain only a
portion of the friction ridge detail that is present on the finger, that is, a ―partial‖ fingerprint.

Fingerprint impressions developed and preserved using any of the above methods can be
digitized by scanning the inked card, lift, item, or photograph. Digital images acquired by
this method are known as ―off-line‖ images. (Typically, the scanners are not designed
specifically for fingerprint applications.)

Since the early 1970s, fingerprint sensors have been built that can acquire a ―livescan‖ digital
fingerprint image directly from a finger without the intermediate use of ink and a paper card.
Although off-line images are still in use in certain forensic and government applications,
on-line fingerprint images are increasingly being used. [16] The main parameters
characterizing a digital fingerprint image are resolution area, number of pixels, geometric
accuracy, contrast, and geometric distortion. CJIS released specifications, known as
Appendix F and Appendix G, that regulate the quality and the format of fingerprint images
and FBI-compliant scanners. All livescan devices manufactured for use in forensic and
government law enforcement applications are FBI compliant. Most of the livescan devices
manufactured to be used in commercial applications, such as computer log-on, do not meet

52
FBI specifications but, on the other hand, are usually more user-friendly, compact, and
significantly less expensive.

Figure 6.2 Live Scan

(a) A good-quality fingerprint; (b) A medium-quality fingerprint with creases; (c) A poor-
quality fingerprint; (d) A very poor-quality fingerprint containing a lot of noise.

Algorithms also typically run on an integrated booking management system to provide real-
time previews (graphical user interface and zoom) to assist the operator in placing or aligning
fingers or palms correctly. Typically, a fingerprint image quality-checking algo-rithm is also
run to alert the operator about the acquisition of a poor-quality fingerprint image so that a
better quality image can be reacquired from the finger or palm. Typical output from such an
automatic quality-checker algorithm is depicted in Figure 6–7.

Swipe sensors, where a user is required to swipe his or her finger across a livescan sensor
that is wide but very short, can offer the lowest cost and size. Such sensors image a single
line or just a few lines (slice) of a fingerprint, and an image-stitching algorithm is used to
stitch the lines or slices to form a two-dimensional fingerprint image (Figure 6–8).

Depending on the application, it may be desirable to implement one or more of the following
algorithms in the live scan device:

53
• Automatic finger-detection algorithm—The scanner automatically keeps looking for the
presence of a finger on its surface and, as soon as it determines that there is a finger
present on its surface, it alerts the system.

• Automatic fingerprint-capture algorithm—Immediately after the system has been alerted

that a finger is present on the surface of the scanner, it starts receiving a series of images,
and the fingerprint-capture algorithm automatically determines which frame in the image
sequence has the best image quality and grabs that frame from the video for further image
processing and matching.

• Vitality detection algorithm—The scanner can determine whether the finger is consistent
with deposition by a living human being.

• Image data-compression algorithm—Compressed image will require less storage and

bandwidth when transferred to the system.

• Image-processing algorithms—Certain applications will benefit from feature extraction

carried out on the sensor itself; the transfer of the fingerprint features will also require less
bandwidth than the image.

• Fingerprint-matching algorithm—Certain applications would like the fingerprint matching

to be performed on the sensor for security reasons, especially for on-board sequence
checking.

• Cryptographic algorithms and protocol(s)—Implemented in the scanner to carry out secure

communication.

6.1.3 Image Enhancement

Fingerprint images originating from different sources may have different noise characteristics
and thus may require some enhancement algorithms based on the type of noise. For example,
latent fingerprint images can contain a variety of artifacts and noise. Inked fingerprints can
contain blobs or broken ridges that are due to an excessive or inadequate amount of ink. Filed
paper cards may contain inscriptions overlapping the fingerprints and so forth. The goal of
fingerprint enhancement algorithms is to produce an image that does not contain artificially
generated ridge structure that might later result in the detection of false minutiae features

54
while capturing the maximum available ridge structure to allow detection of true minutiae.
Adapting the enhancement process to the figure.

Figure 6.3 Finger Image Enhancement

As the user sweeps his or her finger on the sensor, the sensor delivers new image slices,
which are combined into a two-dimensional image.

A fingerprint may contain such poor-quality areas that the local ridge orientation and
frequency estimation algorithms are completely wrong. An enhancement algorithm that can
reliably locate (and mask) these extremely poor-quality areas is very useful for the later
feature detection and individualization stages by preventing false or unreliable features from
being created. [14] Fingerprint images can sometimes be of poor quality because of noise
introduced during the acquisition process. For example: a finger may be dirty, a latent print
may be lifted from a difficult surface, the acquisition medium (paper card or livescan) may
be dirty, or noise may be introduced during the interaction of the finger with the sensing
surface (such as slippage or other inconsistent contact).

6.1.4. Enhancement of Latent Prints for AFIS Searching.

In the case of latent searches into the forensic AFISs, the enhancement algorithm is
interactive, that is, live feedback about the enhancement is provided to the forensic expert
through a graphical user interface. Through this interface, the forensic expert is able to use
various algorithms to choose the region of interest in the fingerprint image, crop the image,
invert color, adjust intensity, flip the image, magnify the image, resize the image window,

55
and apply compression and decompression algorithms. The forensic expert can selectively
apply many of the available enhancement algorithms (or select the parameters of the
algorithm) based on the visual feedback. Such algorithms may include histogram
equalization, image intensity rescaling, image intensity adjustments with high and low
thresholds, local or global contrast enhancement, local or global background subtraction,
sharpness adjustments (applying high-pass filter), background suppression (low-pass filter),
gamma adjustments, brightness and contrast adjustments, and so forth. An example of local
area contrast enhancement is shown in Figure 6–9. In this example, the fingerprint image
enhancement algorithm enhances only a small, square, local area of the image at a time but
traverses over the entire image in a raster scan fashion such that the entire image is enhanced.
Subsequent fingerprint feature extraction can then be either performed manually or through
automatic fingerprint feature extraction algorithms.

Figure 6.4 Enhancement of latent finger prints

An example of local area contrast enhancement. The algorithm enhances the entire image by
enhancing a large number of small square local areas.

6.1.5. Automated Enhancement of Fingerprint Images.

In the case of lights-out applications (frequently used in automated background checks and
commercial applications for control of physical access), human assistance does not occur in
the fingerprint individualization process. Enhancement algorithms are used in the fully
automated mode to improve the fingerprint ridge structures in poor-quality fingerprint

56
images. An example of a fully automated fingerprint image-enhancement algorithm is shown
in Figure 6–10. In this example, contextual filtering is used that has a low-pass (smoothing)
effect along the fingerprint ridges and a band-pass (differentiating) effect in the direction
orthogonal to the ridges to increase the contrast between ridges and valleys. Often, oriented
band-pass filters are used for such filtering. One such type of commonly used filters is known
as Gabor filters. The local context is provided to such contextual filters in terms of local
orientation and local ridge frequency.

6.2 Feature Extraction

Local fingerprint ridge singularities, commonly known as minutiae points, have been
traditionally used by forensic experts as discriminating features in fingerprint images. The
most common local singularities are ridge endings and ridge bifurcations. Other types of
minutiae mentioned in the literature, such as the lake, island, spur, crossover, and so forth
(with the exception of dots), are simply compos-ites of ridge endings and bifurcations.
Composite minutiae, made up of two to four minutiae occurring very close to each other,
have also been used. In manual latent print processing, a forensic expert would visually
locate the mi-nutiae in a fingerprint image and note its location, the ori-entation of the ridge
on which it resides, and the minutiae type. Automatic fingerprint feature-extraction
algorithms were developed to imitate minutiae.

57
 Figure 6.5 contextual filtering-based fingerprint image enhancement algorithms.

location performed by forensic experts. However, most automatic fingerprint minutiae-

extraction algorithms only consider ridge endings and bifurcations because other types of
ridge detail are very difficult to automatically extract. Further, most algorithms do not
differentiate between ridge endings and bifurcations because they can be indistinguishable as
a result of finger pressure differences during acquisition or artifacts introduced during the
application of the enhancement algorithm.

One common approach followed by the fingerprint feature extraction algorithms is to first
use a binarization algorithm to convert the gray-scale-enhanced fingerprint image into binary
(black and white) form, where all black pixels correspond to ridges and all white pixels
correspond to valleys. The binarization algorithm ranges from simple thresholding of the
enhanced image to very sophisticated ridge location algorithms. Thereafter, a thinning
algorithm is used to convert the binary fingerprint image into a single pixel width about the
ridge centerline. The central idea of the thinning process is to perform successive (iterative)
erosions of the outermost layers of a shape until a connected unit-width set of lines (or
skeletons) is obtained. Several algorithms exist for thinning. Additional steps in the thinning

58
algorithm are used to fill pores and eliminate noise that may result in the detection of false
minutiae points.

Figure 6.6 Stages in a typical fingerprint minutiae extraction algorithm.

The resulting image from the thinning algorithm is called a thinned image or skeletal image.
A minutiae detection algorithm is applied to this skeletal image to locate the x and y
coordinates as well as the orientation (theta) of the minutiae points. In the skeletal image, by
definition, all pixels on a ridge have two neighboring pixels in the immediate neighborhood.
If a pixel has only one neighboring pixel, it is determined to be a ridge ending and if a pixel
has three neighboring pixels, it is determined to be a ridge bifurcation.

Each of the algorithms used in fingerprint image enhancement and minutiae extraction has its
own limitation and results in imperfect processing, especially when the input fingerprint
image includes non-friction-ridge noise. As a result, many false minutiae may be detected by
the minutiae detection algorithm. To alleviate this problem, often a minutiae post processing
algorithm is used to confirm or validate the detected minutiae. Only those minutiae that pass
this post processing algorithm are kept and the rest are removed. For example, if a ridge
length running away from the minutia point is sufficient or if the ridge direction at the point
is within acceptable limits, the minutia is kept.

59
The post processing might also include an examination of the local image quality,
neighboring detections, or other indicators of non fingerprint structure in the area. Further,
the image can be inverted in gray scale, converting white to black and black to white.
Reprocessing of this inverted image should yield minutiae endings in place of bifurcations,
and vice versa, allowing a validity check on the previously detected minutiae. The final
detected minutiae are those that meet all of the validity checks. Figure 6–11 shows the steps
in a typical fingerprint feature-extraction algorithm; the extracted minutiae are displayed
overlapping on the input image for visualization.

Many other features are often also extracted in addition to minutiae. These additional features
often provide useful information that can be used in the later matching stages to improve the
fingerprint-matching accuracy. For example, minutiae confidence, ridge counts between
minutiae, ridge count confidence, core and delta locations, local quality measures, and so
forth, can be extracted. These additional features may be useful to achieve added selectivity
from a minutiae-matching process. Their usefulness for this purpose may be mediated by the
confidence associated with each such feature. Therefore, it is important to collect confidence
data as a part of the image-enhancement and feature-extraction process to be able to properly
qualify detected minutiae and associated features.

The early fingerprint feature-extraction algorithms were developed to imitate feature

extraction by forensic experts. Recently, a number of automatic fingerprint feature-extraction
(and matching) algorithms have emerged that use non-minutiae-based information in the

6.3 Matching
Fingerprint matching can be defined as the exercise of finding the similarity or dissimilarity
in any two given fingerprint images. Fingerprint matching can be best visualized by taking a
paper copy of a file fingerprint image with its minutiae marked or overlaid and a
transparency of a search fingerprint with its minutiae marked or overlaid. By placing the
transparency of the search print over the paper copy of the file fingerprint and translating and
rotating the transparency, one can locate the minutiae points that are common in both prints.
From the number of common minutiae found, their closeness of fit, the quality of the
fingerprint images, and any contradictory minutiae matching information, it is possible to
assess the similarity of the two prints. Manual fingerprint matching is a very tedious task.

60
Automatic fingerprint-matching algorithms work on the result of fingerprint feature-
extraction algorithms and find the similarity or dissimilarity in any two given sets of
minutiae. Automatic fingerprint matching can perform fingerprint comparisons at the rate of
tens of thousands of times each second, and the results can be sorted according to the degree

It is important to note, however, that automatic fingerprint-matching algorithms are

significantly less accurate than a well-trained forensic expert. Even so, depending on the
application and the fingerprint image quality, the automatic-fingerprint-matching algorithms
can significantly reduce the work for forensic experts.

Automatic fingerprint-matching algorithms yield imperfect results because of the difficult

problem posed by large intraclass variations (variability in different impressions of the same
finger) present in the fingerprints. These intraclass variations arise from the following factors
that vary during different acquisition of the same finger: (1) displacement, (2) rotation, (3)
partial overlap, (4) nonlinear distortion because of pressing of the elastic three-dimensional
finger onto a rigid two-dimensional imaging surface, (5) pressure, (6) skin conditions, (7)
noise introduced by the imaging environment, and (8) errors introduced by the automatic
feature-extraction algorithms.

Figure 6.7 Stages in a typical fingerprint minutiae matching algorithm.

61
Certain alignment algorithms also take into account the variability caused by nonlinear
distortion. The alignment algorithm must also be able to take into consideration the fact that
the feature extraction algorithm is imperfect and may have introduced false minutiae points
and, at the same time, may have missed detecting some of the genuine minutiae points. Many
fingerprint alignment algorithms exist. Some may use the core and delta points, if extracted,
to align the fingerprints. Others use point pattern-matching algorithms such as Hough
transform (a standard tool in pattern recognition that allows recognition of global patterns in
the feature space by recognition of local patterns in a transformed parameter space),
relaxation, algebraic and operational research solutions, ―tree pruning,‖ energy minimization,
and so forth, to align minutiae points directly. Others use thinned ridge matching or

Once an alignment has been established, the minutiae from the two fingerprints often do not
exactly overlay each other because of the small residual errors in the alignment algorithm and
the nonlinear distortions. The next stage in a fingerprint minutiae-matching algorithm, which
establishes the minutiae in the two sets that are corresponding and those that are
noncorresponding, is based on using some tolerances in the minutiae locations and orienta-
tion to declare a correspondence. Because of noise that is introduced by skin condition,
recording environment, imaging environment, and the imperfection of automatic fingerprint
feature-extraction algorithms, the number of corresponding minutiae is usually found to be
less than the total number of minutiae in either of the minutiae sets in the overlapping area.
So, finally, a score computation algorithm is used to compute a matching score. The
matching score essentially conveys the confidence of the fingerprint matching algorithm and
can be viewed as an indication of the probability that the two fingerprints come from the
same finger. The higher the matching score, the more likely it is that the fingerprints are
mated (and, conversely, the lower the score, the less likely there is a match). There are many
score computation algorithms that are used. They range from simple ones that count the
number of matching minutiae normalized by the total number of minutiae in the two
fingerprints in the overlapping area to very complex probability-theory-based, or statistical-
pattern-recognition-classifier-based algorithms that take into account a number of features
such as the area of overlap, the quality of the fingerprints, residual distances between the
matching minutiae, the quality of individual minutiae, and so forth. Figure 6–12 depicts the
steps in a typical fingerprint matching algorithm. Note that the stages and algorithms
62
described in this sec-tion represent only a typical fingerprint minutiae-matching algorithm.
Many fingerprint minutiae-matching algorithms exist and they all differ from one another. As
with the various extraction algorithms, matching algorithms use different implementations,
different stages, and different orders of stages.

6.3.1 Indexing and Retrieval

In the previous section, the fingerprint matching problem was defined as finding the
similarity in any two given fingerprints. There are many situations, such as controlling
physical access within a location or affirming ownership of a legal document (such as a
driver‘s license), where a single match between two fingerprints may suffice. However, in a
large majority of forensic and government applications, such as latent fingerprint
individualization and background checks, it is required that multiple fingerprints (in fact, up
to 10 fingerprints from the 10 fingers of the same person) be matched against a large number
of fingerprints present in a database. In these applications, a very large amount of fingerprint
searching and matching is needed to be performed for a single individualization. This is very
time-consuming, even for automatic fingerprint-matching algorithms. So it becomes
desirable (although not necessary) to use automatic fingerprint indexing and retrieval
algorithms to make the search faster.

Traditionally, such indexing and retrieval has been performed manually by forensic experts
through indexing of fingerprint paper cards into file cabinets based on fingerprint pattern
classification information as defined by a particular fingerprint classification system.

Similar to the development of the first automatic fingerprint feature extraction and matching
algorithms, the initial automatic fingerprint indexing algorithms were developed to imitate
forensic experts. These algorithms were built to classify fingerprint images into typically five
classes (e.g., left loop, right loop, whorl, arch, and tented arch) based on the many fingerprint
features automatically extracted from fingerprint images. (Many algorithms used only four
classes because arch and tented arch types are often difficult to distinguish.)

Fingerprint pattern classification can be determined by explicitly characterizing regions of a

fingerprint as belonging to a particular shape or through implementation of one of many
possible generalized classifiers (e.g., neural networks) trained to recognize the specified

63
patterns. The singular shapes (e.g., cores and deltas) in a fingerprint image are typically
detected using algorithms based on the fingerprint orientation image. The explicit (rule-
based) fingerprint classification systems first detect the fingerprint singularities (cores and
deltas) and then apply a set of rules (e.g., arches and tented arches often have no cores; loops
have one core and one delta; whorls have two cores and two deltas) to determine the pattern
type of the fingerprint image (Figure). The most successful generalized (e.g., neural network-
based) fingerprint classification systems use a combination of several different classifiers.

Many of the newer automatic fingerprint classification algorithms do not use explicit classes
of fingerprints in distinct classifications but rather use a continuous classification of
fingerprints that is not intuitive for manual processing but is amenable to automatic search
algorithms. In continuous classification, fingerprints are associated with numerical vectors
summarizing their main features. These feature vectors are created through a similarity-
preserving transformation, so that similar fingerprints are mapped into close points (vectors)
in the multidimensional space. The retrieval is performed by matching the input fingerprint
with those in the database whose corresponding vectors are close to the searched one. Spatial
data structures can be used for indexing very large databases. A continuous classification
approach allows the problem of exclusive membership of ambiguous fingerprints to be
avoided and the system‘s efficiency and accuracy to be balanced by adjusting the size of the
neighborhood considered.

In general, different strategies may be defined for the same indexing mechanism. For
instance, the search may be stopped when a fixed portion of the database has been explored
or as soon as a matching fingerprint is found. (In latent fingerprint individualization, a
forensic expert visually examines the fingerprints that are considered sufficiently similar by
the minutiae matcher and terminates the search when a true correspondence is found.) If an
exclusive classification technique is used for indexing, the following retrieval strategies can
be used:

• Hypothesized class only—Only fingerprints belonging to the class to which the input
fingerprint has been assigned are retrieved.

64
 Figure 6.8 six commonly used fingerprint classes

The six commonly used fingerprint classes: (a) whorl, (b) right loop, (c) arch, (d) tented arch,
(e) left loop, and (f) double loop whorl.
• Fixed search order—The search continues until a match is found or the whole database has
been explored. If a correspondence is not found within the hypothesized class, the search
continues in another class, and so on.

• Variable search order—The different classes are visited according to the class likelihoods
produced by the classifier for the input fingerprint. The search may be stopped as soon as a
match is found or when the likelihood ratio between the current class and the next to be
visited is less than a fixed threshold.

Finally, many system-level design choices may also be used to make the retrieval fast. For
example, the search can be spread across many computers, and special-purpose hardware
accelerators may be used to conduct fast fingerprint matching against a large database.
may be better than algorithm B at a low false-positive rate, but algorithm B may be better
than algorithm [16] A at a low false non-match rate. In such cases, the algorithm designers
may choose a certain algorithm or specific parameters to be used, depending on the
application.

65
6.4 Fingerprint Standards
Law enforcement agencies around the world have had standards for the local exchange of
inked fingerprints for decades. In 1995, Interpol held a meeting to address the transfer of ink-
and-paper fingerprint cards (also known as forms) between countries. The local standards
naturally had different text fields, had different layouts of text fields, were in different
languages, and were on many different sizes of paper. Before that effort could lead to an
internationally accepted fingerprint form, Interpol moved to the electronic exchange of
fingerprints. In the ink-and-paper era, the standards included fiber content and thickness of the
paper, durability of the ink, size of the ―finger boxes‖, and so forth. With the move in the early 1990s
toward near real-time responses to criminal fingerprint submittals, there came a new set of standards.

The only way to submit, search, and determine the status of fingerprints in a few hours from
a remote site is through electronic submittal and electronic responses. The source can still be
ink-and-paper, but the images need to be digitized and submitted electronically to address the
growing demand for rapid turnaround of fingerprint transactions.

The FBI was the first agency to move to large-scale electronic submission of fingerprints
from remote sites. As part of the development of IAFIS, the FBI worked very closely with
NIST to develop appropriate standards for the electronic transmission of fingerprint images.

Starting in 1991, NIST held a series of workshops with forensic experts, fingerprint
repository managers, industry representatives, and consultants to develop a standard, under
the ANSI guidelines, for the exchange of fingerprint images. It was approved in November
1993, and the formal title was ―Data Format for the Interchange of Fingerprint Information
(ANSI NIST-CSL 1-1993)‖. This standard was based on the 1986 ANSI/National Bureau of
Standards minutiae-based standard and ANSI/NBS-ICST 1-1986, a standard that did not
address image files.

This 1993 NIST standard (and the later revisions) became known in the fingerprint
technology world simply as the ―ANSI/NIST standard‖. [13] If implemented correctly (i.e.,
in full compliance with the standard and the FBI‘s implementation), it would permit
fingerprints collected on a compliant livescan from any vendor to be read by any other
compliant AFIS and the FBI‘s yet-to-be-built (at that time) IAFIS.

66
The standard was deliberately open to permit communities of users (also known as domains
of interest) to customize it to meet their needs. Some of the customizable areas were image
density (8-bit gray scale or binary) and text fields associated with a transaction (e.g., name,
crime). The idea was that different communities of users would write their own
implementation plans. The mandatory parts of the ANSI/NIST standard were the definitions
of the record types, the binary formats for fingerprint and signature images and, within
certain record types, the definition of ―header‖ fields such as image compression type.

6.4.1 Record Types.

For a transaction to be considered ANSI/NIST compliant, the data must be sent in a
structured fashion with a series of records that align with ANSI/NIST record types as
implemented in a specific user domain (e.g., Interpol).

• All transmissions (also known as transactions) have to start with a type 1 record that is
basically a table of contents for the transmission, the transaction type field (e.g., CAR for
―criminal tenprint submission—answer required‖), and the identity of both the sending and
receiving agencies.
• Type 2 records can contain user-defined information associated with the subject of the
fingerprint transmission (such as name, date of birth, etc.) and the purpose of the transaction
(arrest cycle, applicant background check, etc.). These fields are defined in the domain-of-
interest implementation standard (e.g., the FBI‘s EFTS). Note that type 2 records are also
used for responses from AFISs. They fall into two sets: error messages and search results.
The actual use is defined in the domain specification.

• Types 3 (low-resolution gray scale), 4 (high-resolution gray scale), 5 (low-resolution

binary), and 6 (high-resolution binary) were set up for the transmission of fingerprint images
at different standards (500 ppi for high resolution and 256 ppi for low resolution) and image
density (8 bits per pixel for grayscale) or binary (1 bit per pixel for black and white). Note
that all images for records type 3 through 6 are to be acquired at a minimum of 500 ppi;
however, low-resolution images are down-sampled to 256 ppi for transmission. There are
few, if any, ANSI/NIST implementations that support type 3, 5, or 6 images (see explanation
below). None of these three record types are recommended for use by latent examiners and
fingerprint technicians.
67
• Type 7 was established for user-defined images (e.g., latent images, faces) and, until the
update of the ANSI/ NIST standard in 2000, it was the record type for exchanging latent
images. This record type can be used to send scanned copies of identity documents, and so
forth. Again, the domain specification determines the legitimate uses of the type 7 record.

• Type 8 was defined for signatures (of the subject or person taking the fingerprints), and it is
not used in many domains.

• Type 9 was defined for a minimal set of minutiae that could be sent to any AFIS that was
ANSI/NIST-compliant.

6.4.2 Finger Image Quality.

Both the ANSI/NIST standard and the EFTS lacked any metrics or standards for image qual-
ity. The FBI then appended the EFTS with an image quality standard (IQS) known as
Appendix F. (Later, a reduced set of image quality specifications were added as Appendix G
because the industry was not uniformly ready to meet Appendix F standards.) The IQS
defines minimal acceptable standards for the equipment used to capture the fingerprints.
There are six engineering terms specified in the IQS. They are:

1. Geometric image accuracy—the ability of the scanner to keep relative distances between
points on an object (e.g., two minutiae) the same relative distances apart in the output
image.

2. Modulation transfer function (MTF)—the ability of the scanning device to capture both
low-frequency (ridges themselves) and high-frequency (ridge edge details) information in
a fingerprint at minimum standards.

3. Signal-to-noise ratio—the ability of the scanning device to digitize the information without
introducing too much electronic noise (that is, with the pure white image parts appearing
pure white and the totally black image parts appearing totally black).

4. Gray-scale range of image data—avoiding excessively low-contrast images by ensuring

that the image data are spread across a minimal number of shades of gray.

5. Gray-scale linearity—as the level of gray changes in a fingerprint capture, the digital
image reflects a corresponding ratio of gray level across all shades of gray.

68
6. Output gray-level uniformity—the ability of the scanning device to create an image with a
continuous gray scale across an area on the input image (tested using a special test image)
that has a single gray level.

Interestingly, only two of these six image quality standards apply to latent scanning devices:
geometric image accuracy and MTF. [15] In fact, the FBI does not certify (see below for a
discussion of certified products) scanners for latent use but recommends that latent examiners
purchase equipment they are comfortable with using from an image-quality perspective. But
EFTS Appendix F does mandate that latent images be captured at 1,000 ppi.

There are no standards for the quality of the actual fingerprint, but livescan and AFIS
vendors have rated fingerprint quality for years. They know that fingerprint quality is
possibly the strongest factor in the reliability of an AFIS‘s successfully matching a
fingerprint to one in the repository. These ratings are often factored into the AFIS algorithms.

Although no standard exists for fingerprint image quality, NIST has researched the
relationship between calculated image quality (using algorithms similar to those employed by
AFIS vendors) and successful match rates in automated fingerprint identification systems.
This led NIST to develop and publish a software utility to measure fingerprint image quality.

6.5 Latent Interoperability

When IAFIS was being developed, the FBI established (in the EFTS) two ways for latent
impressions to be run through IAFIS from outside agencies.

Remote Submittals. The agency with the latent impression can send (electronically or via
the mail) the impression (as an image in the case of electronic submittal) to the FBI, and FBI
staff will perform the editing, encoding, searching, and candidate evaluation. The FBI will
make any identification decision and return the results to the submitting agency. This process
mimics the pre-IAFIS workflow but adds the option of electronic submittal.

Remote Searches. The agency with the latent impression performs the editing and encoding
and then sends (electronically) a latent fingerprint features search (LFFS) to IAFIS for lights-
out searching. IAFIS then returns a candidate list, including finger images, to the originating
agency to perform candidate evaluation. The submitting agency makes any identification

69
decision. To support LFFS remote search capability, the FBI published the ―native‖ IAFIS
feature set definition.

6.6 Summary of this Chapter

Because most AFISs (other than IAFIS) do not have remote LFFS functionality (as of 2007),
latent interoperability at the image level usually requires labor on the part of the searching
agency. The desire to move that labor burden to the submitting agency is natural because
many have some level of excess capacity that could possibly support remote latent searches
during off-hours.

70
 CHAPTER 7
BIOMETRIC SECURITY USING FINGER
PRINT RECOGNITION
7.1 Biometric Security Using Finger Print Recognition
The convenience of current electronic applications has led to an explosive increase in their
use. E-banking, electronic fund transfer, online shopping and virtual auctions are just some
applications prevalently used by the public. Trust, as a result, has become more of an issue.
As expected, there is an element of security when performing transactions in person. When
using electronic services, however, this element is removed. In an electronic environment, it
is easy for a person to pose as someone else. As a result, the verification of an individual‘s
identity is vitally important. [15] No one wants automatic teller machines (ATMs) or credit
card companies to authorize fraudulent transactions made by someone else. To try to prevent
illegal transactions, electronic security methods are employed. Most methods rely on one of
two basic security methodologies: token based security and secret-based security. Token-
based security relies on the user‘s possession of a special item or token. Often, this token is
an access or ID card such as a credit card or security access badge. Secret-based security, on
the other hand, relies on an individual‘s secret ID number like a computer password or a PIN
number that only that person would know. This information is then supplied to verify his or
her identity. Both methodologies have one major flaw. Neither can accurately determine if
the individual that possesses a token or knows some secret information is actually the
individual it represents. Tokens can be stolen, and secret information can be guessed or
fraudulently obtained. Once a person has the token or secret password, it is easy to pose as
the original owner. A third security methodology that seeks to solve the problem of positive
identification is biometrics. Biometrics is the practice of using people‘s physical or
behavioral traits to confirm their identities. People have been using biometrics for a long time
without even realizing it. Most people identify others based upon physical traits such as their
body shapes, facial features or voices. This is the same principle behind modern biometric
identification. Physical traits such as facial structure, retina pattern, hand geometry, iris
pattern and fingerprints are all used by biometric systems to try to verify an individual‘s
identity. One trait of particular interest, since it is easily collected and universal is
fingerprints. The law enforcement community has been using fingerprint identification for
many years. Although a segment of the public views fingerprinting as an invasion of privacy,
most people do not. As a result, fingerprint information promises to be a common method of
authentication in the future. For this to occur, however, an efficient and cost effective method

72
for obtaining fingerprint images must be developed. The three phases The typical biometric
system has three distinct phases. These are biometric data acquisition, feature extraction and
decision making. The first step, data acquisition, is the collection of the actual biometric
information. In this case, it would be the individual‘s fingerprint. The acquisition phase is
extremely important. If high quality images are not obtained, the next phases cannot operate
reliably. The second phase is feature extraction. In this phase, important information is
extracted from the image supplied by the acquisition phase. This data is typically a pattern of
features or landmarks that allows a given individual to be uniquely identified. For fingerprint
recognition, these features are typically minutia points such as ridge endings (the termination
points of fingerprint friction ridges) and ridge bifurcations. (The points where a ridge splits to
form a Y shape.) Figure 1 shows examples of ridge endings and bifurcations. The coordinate
pattern of these features is unique to a given individual; thus, it can serve as his or her
identification. [14] The final phase of a biometric system is the decision-making phase. In
this phase, the feature pattern that was extracted from the image is compared to a previously
known example. A decision is then made regarding the identity of the individual. If the
patterns are close enough, the individual‘s identity is verified; otherwise, the claimed identity
is rejected. So if an accurate image is not obtained in the first phase, the rest of the process
will be inaccurate. In fact, most difficulties in accurately identifying an individual can be
traced back to the image acquisition phase. As a result, the improvement of these acquisition
systems is a key area of research.

7.1.1 Problems in collecting fingerprints

There are five major problems in current fingerprint acquisition technology. The first
challenge is inconsistent contact. Most fingerprint collection systems require that the finger
come into contact with a surface. This is often a glass platen or another kind of sensing plate.
By placing the finger on a surface, the three dimensional finger is being mapped onto a two
dimensional plane. As a result distortions occur. The skin on the finger will stretch when
pressed against the surface. This stretching causes slight changes in the distances between
features on the fingerprint. In most cases, different levels of pressure will be used each time
the individual‘s finger is scanned. This results in slightly different images as well. The
second challenge is non-uniform contact. The non-uniform contact occurs due to minor

73
 Figure 7.1 problems in collecting fingerprints

inconsistencies in the surface of the finger. For a fingerprint-scanning device to capture the
whole fingerprint, it is necessary for the whole print to be in contact with the scanning
surface. But ridges worn down or dry in comparison to surrounding ridges will often not
come in contact with the scanning surface. Other factors that can interfere include sweat, dirt,
grease and skin disease. Irreproducible contact is another major problem for fingerprint
acquisition systems. Irreproducible contact occurs when the captured image changes from
session to session. For example, an individual could have their fingerprint altered through
accidental injury. Cuts, scratches, scars and ridges that are worn down all contribute to
changes in a fingerprint image. These changes cause problems for the recognition algorithms.
When dealing with electronic sensors and scanners, in many cases different parts of the
fingerprint are imaged during each session. Most scanners are not large enough to capture the
entire surface of the finger. As a result, only a piece of the total fingerprint is captured. By
imaging different portions of the finger each time, additional minutia features are inserted
and some previously measured features are removed. This change from session to session is
an additional problem for identification algorithms. Another challenge is noise. Noise can be
introduced into the system through the acquisition process. This noise can be caused by
electromagnetic radiation, excessive ambient light, or imperfections in the scanning
equipment. Another type of noise that can be introduced is residual noise. This is noise that
enters that image due to residual portions of previous prints. For example, an impression of
the previous fingerprint is often left behind on the scanner surface. The impression, usually
an oily residue, can interfere with the next scanned fingerprint. The previous impression may
just add excess grease to the next scan, or it may result in a ghost image of the previous
fingerprint. In either case, the unwanted information causes serious problems. The final
major problem is with the feature extraction system. In many cases, some of the signal
processing algorithms used to enhance an acquired image can leave irregularities in the

74
image causing problems for the feature extraction stage. These signal-processing algorithms
may be implemented in software as part of the feature extraction stage, or in hardware as part
of the acquisition system. In either case, the irregularities that they introduce cause additional
problems when trying to find a match to a given print.

7.1.2 Ink-based fingerprinting

The oldest and most common method for obtaining fingerprints is using ink and paper. It has
been used by law enforcement agencies for years and is probably the method most people
think of when the topic is mentioned. The process of performing ink and paper-based
fingerprinting is relatively simple. First, the finger must be evenly coated with a layer of ink
done by rolling the finger on an ink-covered surface. Once the finger is coated with ink, there
are two methods that are typically used to acquire an image of the print. The first and most
common method is called ―rolling.‖ [12] To obtain a print using the rolling method, the ink-
coated finger is rolled on a piece of paper starting with the edge of the fingernail on one side
and continuing to the edge of the fingernail on the other. Figure 2 shows a fingerprint
acquired using the rolling technique. This process provides an impression of a large portion
of the finger surface. Having such a large portion of the finger surface available is a major
benefit of this particular method. The larger area allows more useful information to be
gathered from the print. A larger number of usable minutia points in the image allow for
easier matching with future copies of the fingerprint. This is especially useful when the later
print captures may only contain a portion of the finger surface. Unfortunately, the act of
rolling the finger tends to cause distortion in the resulting image. The pressure applied to the
finger stretches the skin causing small distance changes. The second method is called
―dabbing.‖ As the name suggests, the dabbing method involves simply pressing the inked
finger onto the paper. Since the finger is not rolled, less of the fingerprint surface is captured.
However, the dabbing method causes less distortion in the fingerprint image because
stretching is minimized. When fingerprinting is performed by law enforcement agencies, all
ten fingers are typically done. The resulting images are placed on cards and stored for later
identification. In the past, law enforcement agencies needed to perform time consuming
visual comparisons between an unknown print and sets of prints stored in their card database.
In recent years, however, many agencies have switched to more computerized systems. The

75
cards containing the fingerprints are scanned into a computer using a flatbed scanner or CCD
camera. Once in the computer, the number of possible matches can be limited by applying
special processing and match-searching algorithms. This limited set of possible matches can
then be searched manually in much less time. Even with computer matching systems, ink and
paper-based fingerprinting is not a very practical solution. Since the inked print must be
scanned into the computer, the time required to verify would prohibit timely authentication.
In addition, the requirement of placing ink on the fingers is not typically acceptable for an
application that requires frequent sampling of the fingerprint. Not many people would want
to get their fingers inked every time they want to log into their computer or

7.1.3 Optical methods

The remaining technologies described can all be grouped into a category called ―live scan.‖
This term means that the fingerprint images are captured by the computer directly from the
individual‘s finger. These technologies typically capture a smaller portion of the finger
surface than ink and paper-based methods, but they are much quicker and do not have
unwanted side effects like ink-covered fingers. The most common technique used by optical
fingerprint scanners is that of frustrated total internal reflection. The typical scanner design
consists of a glass platen or prism, a light source and a light detector. The frustrated total
internal reflection technique works by using some basic principles of optics. When a finger is
placed on the glass plate or prism, the ridges of the fingerprint come into direct contact with
the glass material. The valleys, however, do not directly touch the surface. The behavior of
the glass plate relative to the incident light is changed due to the contact with the ridges.
When no ridge is present, the light that hits the glass plate at an angle is reflected back where
it strikes the sensor unit. The light that strikes the portions of the platen where a ridge is
present diffracts and the ridges appear black. Figure 3 demonstrates the basic operation of
frustrated total internal reflection. The benefits are that the scanners are inexpensive to
manufacture and are capable of relatively high resolutions. On the other hand, optical
scanners are very sensitive to ambient light. If there is too much light in the area where the
scanner is being used, it can lead to poor image quality since stray light can confuse the

76
 Figure 7.2 Optical method of fingerprint

detector. In addition, optical fingerprint scanners tend to suffer from a lack of contrast. The
images can be extremely light which makes them more difficult to process. In many cases,
additional signal processing is required to darken the images so that accurate feature
extraction can occur. The signal processing algorithms often serve to introduce image
artifacts and amplify the noise in the image. There are two types of optical scanners that are
less often utilized, but show a great deal of promise. The first one is based on hologram
reflection. These devices are similar to those using frustrated total internal reflection, except
that the light passes through a light diffractive grating contained within a hologram. The light
is then modulated by the ridge structure of the finger. The modulated light then returns
through the diffractive grating and is analyzed by a detector. Scanners based on this
technology can overcome some of the distortions introduced by the detector arrays within
standard optical scanners. They can also be built within much smaller packages. This would
make them practical for inclusion within portable devices. The final type of optical scanner
discussed here is a scanner that uses a CCD camera to capture the fingerprint. The finger
does not have to be pressed against a platen. Systems utilizing this no contact technology
usually consist of a CCD camera that has been focused on a specific point. A guide, or finger

77
rest, is typically placed so that the user will position their finger at the correct photographic
point. The CCD camera captures an image of the target finger. This image is then processed
to enhance the ridge structure of the finger. There are several benefits to this type of design.
First, there is no distortion due to skin stretching from pressing against a surface. This helps
to make successive images more reproducible. Also, there is no platen that can be scratched.
This lowers the required system maintenance. Still, high ambient light is likely to be a
problem. In addition, even these optical scanners are still limited to scanning the surface
layer of the skin, which may be worn and unusable.

7.1.4 Thermal imaging

Thermal imaging uses body heat to create an image of the fingerprint. A pyro electric
substance is used within a Fig. 4 A depiction of the interaction between the ridges and
valleys of a fingerprint and the thermal sensors in a thermal fingerprint scanner. A higher
temperature is sensed where the ridge directly touches the surface, and a lower temperature is
detected in the valleys where there is an air gap.

Figure 7.3 Thermal Imaging

Fig. 5 An illustration of a typical electromagnetic fingerprint sensing device. The sensor

plates are close to the finger at the ridges, but separated at the valleys. This causes different

78
capacitance values for the different sensor plates. 36 IEEE POTENTIALS sensor to create an
electrical signal that relies on the heat applied to the sensor. An array of these sensors is
employed to detect the temperature of the finger at various positions. When a finger is placed
on the sensor array, the ridges of the fingerprint come into direct contact with the sensor
surface. Since the ridges touch the sensor and the valleys do not, the ridges will be sensed at
a higher temperature than the valleys. These sensed temperature values are then converted
into an eight-bit gray scale image. The number and density of the sensors within the array
determines the resolution of the resulting image. Figure 4 displays the interaction between
the ridge and valley pattern of a fingerprint and the thermal sensor array of a thermal
fingerprint scanner. Thermal scanners have some benefits over the others. Thermal sensors
are less influenced by the condition of the finger. Whether the finger is wet, dry or greasy
makes very little difference to the scanner. The primary concern of the scanner is that the
temperature variations are sufficient. Therefore, as long as the ridge structure is well defined,
a good image can be obtained. (However, it does not solve the problem of dirty or worn
fingerprints.) In addition to its tolerance to finger conditions, thermal scanners are not
affected by ambient light, and they are often capable of obtaining an image through a thin
film such as a latex glove. As long as the heat pattern of the fingerprint is preserved through
the film, a thermal scanner can obtain an image. The major problem with a thermal scanner is
obviously its sensitivity to environmental heat. The best images are acquired when the
scanner and the finger are at very different temperatures. If the scanner is in an environment
that could cause its temperature to approach that of the finger, the thermal scanner is not a
good choice. In addition, if the ridge pattern is not distinct enough, a thermal scanner cannot
determine the difference between the temperature of a ridge and the temperature of a valley.
This will result in poor image quality.

7.1.5 Electromagnetic field imaging

Fingerprint scanning based on electromagnetic fields is a relatively new area. Most
electromagnetic field sensors are based on the principles behind the capacitor. A typical
capacitor consists of two conducting plates separated by an insulating dielectric material. The
capacitance of the capacitor is based on the charge on the plates, the size of the plates, and
the distance by which the two plates are separated. [13] In a typical electromagnetic

79
fingerprint sensor, the sensor‘s surface consists of a large number of small conductive plates.
The finger is then used as the second plate in the capacitor arrangement. When the finger is
placed on the surface, it comes into contact with a conduction ring that surrounds the sensor
array. This conduction ring passes a charge onto the finger. This charge is propagated by the
highly conductive layer of newly formed cells beneath the outer layer of skin. The scanner
then senses the capacitance on each of the tiny plates within the sensor array and converts
this value into an eight-bit gray scale value. The values generated by each of the small plates
in the array are collected to form an image of the fingerprint. The reason that this works is
actually quite simple. The finger acts as the second plate of the capacitor. It is charged by the
conductive ring. The distance of the finger from the sensing plates is determined by the shape
of the fingerprint. The ridges will be much closer to the sensor plates and, thus, will yield a
capacitance that is reflective of two plates separated by a small distance. The charge at the
valleys, however, is separated from the sensing plates by a much larger distance and will
yield a capacitance that reflects this fact. Figure 5 depicts the arrangement of an
electromagnetic scanning device. One major advantage to using electromagnetic scanning is
that it will return an image of the cells below the outer layer of skin. As a result,
electromagnetic scanning is far less influenced by surface conditions and worn finger surface
ridges. Another advantage of electromagnetic sensors is their size. Since they are solid state
devices, they can be made very small. Typical electromagnetic sensors are about the size of a
postage stamp. This feature makes them ideal for inclusion in portable devices. A weakness
of electromagnetic scanners is their susceptibility to damage from electrostatic discharge.
Since the finger is acting as a plate in the capacitor, the electronics in the scanning device
must be developed so that it can come into contact with the finger. This makes the chip very
open to problems from static electricity that tends to build up on the human body. A
sufficient discharge could destroy an electromagnetic sensor. Developing a design that would
ground the individual before allowing them to touch the chip would solve this problem.

7.1.6 Ultrasound imaging

Ultrasound involves, as the name suggests, the use of sound waves generated at frequencies
higher than the human ear can hear. Many people are probably familiar with the use of
ultrasound in medical applications. It has a proven track record in obtaining images of the

80
human body without the ill effects of radiation. Ultrasound fingerprint scanning is simply an
extension of this proven medical technique. An ultrasound scanner requires that the user
place his or her finger on a glass plate. Once the finger is in place, a transmitter, and ring of
receivers, is mechanically moved along the length of the glass plate. As the assembly is
moved, the transmitter emits pulses of sound. As this sound comes into contact with objects,
some of the sound waves are transmitted through the object and some are reflected back.

Figure 7.4 Ultrasound Imaging

Figure 6 shows the transmission and reflection concept of an ultrasound scanner. The waves
reflected back are received by the ring of receivers. These echoes, and the measured delays in
their return, are then transformed into an image using signal-processing techniques used in
medical computer tomography. Ultrasound imaging obtains images from the layers of skin
slightly beneath the surface of the finger. As a result, much like electromagnetic imaging,
ultrasound imaging is less susceptible to the finger‘s surface conditions. It is also capable of
imaging through items covering the finger, such as latex gloves. These benefits make
ultrasound scanning attractive. The main drawbacks of ultrasonic scanning are speed, size
and cost. Ultrasound scanners require mechanical movement of the transmitter and receiver
assemblies. As a result the scanner package must include all of the necessary mechanical
equipment. In addition, the mechanical movement is slow in comparison with the time
required for the other scanners, which is often measured in milliseconds. Finally, the cost of
the components is high making the overall imaging cost much higher than other types. Thus,
it is unlikely that ultrasound scanners will be seen in portable devices anytime soon.
81
However, they do seem to be an attractive scanner when high accuracy and security are
requirements that outweigh speed and size. sound waves are transmitted through the object
and some are reflected back. Figure 6 shows the transmission and reflection concept of an
ultrasound scanner. The waves reflected back are received by the ring of receivers. These
echoes, and the measured delays in their return, are then transformed into an image using
signal-processing techniques used in medical computer tomography. Ultrasound imaging
obtains images from the layers of skin slightly beneath the surface of the finger. As a result,
much like electromagnetic imaging, ultrasound imaging is less susceptible to the finger‘s
surface conditions. It is also capable of imaging through items covering the finger, such as
latex gloves. These benefits make ultrasound scanning attractive. The main drawbacks of
ultrasonic scanning are speed, size and cost.

7.1.7 Comparison testing

Now let‘s try comparing the performance of these technologies by testing some actual units.
The technologies involved are optical, thermal and ultrasound. The optical scanner that was
used was the Finger Imaging System manufactured by Polaroid. The thermal scanner was the
Sweep manufactured by Thomson CSF, and the ultrasound scanner was a series 500 scanner
from Ultra-Scan, Inc. The same finger was scanned on each of the three scanners under five
different finger conditions.

Figure 7.5 Image Comparison

82
Figures 7 through 9 display the results of the scans. Table 1 displays a ranking of the quality
of the images in each case. These conditions consisted of the following: 1. Normal—no
special action was taken 2. Wet—The finger was moistened with a wet rag before scanning.
3. Dry—The finger was dried out thoroughly. 4. Dirty—The finger was rubbed in soil and
then wiped with a dry cloth before scanning. 5. Gloved—The finger was placed in a latex
glove and then scanned. As can be seen from the results, each scanner performed close to
what would be expected based upon its described strengths and weaknesses. By looking at
the results, a few interesting comments can be made. First in the case of the optical scanner,
the resulting images displayed in this article were contrast-enhanced after scanning so that
they would show up in printing. All the images from the optical scanner were very light and
difficult to see before enhancement. The optical scanner also had a great deal of trouble with
the dirty finger. Finally, there is a ghost image in the glove scan. [12] The optical scanner
should not have any results for a scan through a glove. However, at the time of scanning
there was an oily residual print left on the scanner from a previous scan. The presence, of this
residue resulted in the ghost image, seen in Fig. 7(e). With the thermal scanner, we see that it
can actually sense a print through a glove. However, the resulting print is not of the same
quality attained in the other scans. Some distortion can also be noted in all of the thermal
scans due to the fact that it is a sweep-style scanner. In other words, the image is dependent
upon the sweeping of the finger across a sensing strip. Changes in speed during the scanning
causes the horizontal distortions that can be seen in the results.

7.2 Summary of this chapter

some trouble with the dirty finger. While there was trouble with the dirty finger, the quality
of the scan was higher than that of the optical scanner in the same situation. Finally, the
ultrasound scanner performed relatively well. The scans are of high quality in most of the
cases. It had slight problems with the dirty finger. While the image is not of the highest
quality, it would most likely still be of high enough quality to identify the individual.

83
 CHAPTER 8
PROPOSED SYSTEM FINGERPRINT FOR
ATM
8.1 Fingerprint For ATM system
The Modern banking technology has altered the way banking activities are usually done. One
banking technology that has impacted to banking activities is the automated teller machine
(ATM). Due to ATM technology, a customer is able to perform different banking activities
such as cash withdrawal, transferring money, paying phone bills and electricity bills. In a
short, ATM provides facilities to customers such as quick, easy and convenient way to access
their bank accounts and to conduct financial transactions. Talking about ATM security,
Personal identification number (PIN) or password is very important. PIN or password is
widely used to secure financial/confidential information of customers from unauthorized
access.

Figure 8.1 Fingerprint or ATM System

An ATM is a IT enabled Electro-mechanical system that has connectivity to the accounts of a

banking system. It is computerized machine developed to deliver cash to bank customers
without human intervention; it can be used to transfer money between different bank
accounts and provide basic financial facilities such as balance enquiry, mini instatement, cash
withdrawal, fast cash, etc. In this project the fingerprint sensor sense the thumb impression of
the corresponding person and that image will be compared with registered image, if the both
images are unique, then the finger print device activates particular task like access to the
system, identification of the customer. The project operation contains 2 modes, the first one
is Administration mode and the second is User mode. The Administration mode is used to
register the new user and gives the mode of authorization. The Administration mode has the
ability to create and delete the users. The user mode is mode used for the authentication of
the bank customer. In user mode of authorization, creation and deletion of a user cannot be

85
performed. The paper is arranged as follows. Section II deals with Research background.
Section III provides the key components of Proposed ATM system. Section IV deals with
Hardware Architecture. Section V described Software design. Section VI is dedicated for
Fingerprint Recognition process. Section VII elaborate about GSM technology used. Section
VIII presented the results and the discussions on the results and conclusion.

8.2 Proposed Fingerprint Authenticated ATM System

The architecture of the proposed system is presented in Figure 2. The Internet serves as the
operational environment and platform for the system. The thumbprint database of customers
is available on the Internet and it takes relational database model for stored information on
the thumbprints of all registered customers. These information include type, pattern and
feature characteristics. The flowchart of the thumbprint verification component used for
verifying the authenticity of the user is presented in Figure 3. User verification involves
enrolment, enhancement, feature extraction and matching as shown in Figure 4 [9]. During
image enhancement, the foreground regions of the image which are the regions containing
the ridges and valleys are separated from the background regions, which consist mostly the
noises.

Figure 8.2 Fingerprint Authenticated ATM System

86
Segmentation is performed with a view to ensuring that focus is only on the foreground
regions while the background regions are ignored. Normalization is performed on the
segmented fingerprint image ridge structure so as to standardize the level of variations in the
image grey-level values. By normalization, the grey-level values are brought to a range that
is good enough for improved image contrast and brightness. The normalized image is filtered
for removal of noise and other spurious features. Filtering also preserves the true ridge and
valley structures and it involves ridge orientation and frequency estimations. Filtered image
is binaries and thinned for satisfactory feature extraction. At the feature extraction stage,
major features; namely ridge ending and bifurcation are located and extracted from the image
[12]. These two features are the main characteristics that establish uniqueness among
different fingerprints. The extracted features form a template that is matched with templates
of the other images in the database. The algorithms proposed in [9,10] were implemented for
thumbprint image enhancement in the proposed fingerprint authenticated ATM system.
Similarly, the feature extraction algorithm proposed by I awoken [9] and

8.3 For a customer bifurcation point:

Examine the 3 × 3 neighborhood of the bifurcation point in a clockwise direction. For the
three pixels that are connected with the bifurcation point, label them with the value of 1.
Label with 1 the three ridge pixels that are connected to these three connected pixels. Count
in a clockwise direction, the number of transitions from 0 to 1 (T01) along the border of
image M. If T01=3, then the candidate minutiae point is validated as a true bifurcation For a
candidate ridge ending point: Label with a value of 1 all the pixels in M, which are in the 3 ×
3 neighborhood of the ridge ending point. Count in a clockwise direction, the number of 0 to
1 transitions (T01) along the border of image M. If T01=1, then the candidate minutiae point
is validated as a true ridge ending. The feature matching algorithm involves the following
[16]: The distance, λi between the ith minutiae point Pi (ai ,bi ) and the image core point
M(ρ,σ) is obtained from: 2 2 0.5 λρ ρ i =− − (( ) +( ) ) a b i i (5)

The degree of closeness Ec is obtained for images K and L by using the formula: 1 1 (| ( ) ( )
|)*{ ( )}− = = − ∑ s c i E Gi Hi Gi (6) s is the smaller of the respective number of feature
points in the two images, G(i) and H(i) represent the distance between the ith minutiae point

87
and the core point in K and L respectively. The correlation coefficient value, S between K
and L, is then computed as the pattern matching score by using the formula: 2 (1 )*10− S E =
− c (7) In this case, the closeness value will be Ec =0 for exact or same images and,
consequently, the correlation will be S=1. Decision is taken based on the result of matching.
If there is a match for the thumb being presented, the system proceeds to the ATM
transaction otherwise the whole process is repeated or cancelled. The system is designed to
run on Windows and all other operating system with .Net support. The system sits on top of
the Microsoft SQL (MSSQL) Server as the backend while C Sharp (C#) serves as frontend.
The .NET platform is a new development platform developed and designed to facilitate
internet development and provide a fresh application programming interface (API) to the
services of classic Windows Operating Systems. It also brings together a number of disparate
technologies that emerged from Microsoft.

Figure 8.3 For a customer bifurcation point

The three commonest classes of .NET Framework are Common Language Runtime (CLR),
Base Class Library (BCL) and Top Level Development Target (TLDT). CLR is a managed
execution environment that handles memory allocation, error trapping and interacting with
the operating system services while BCL is an extensive collection of programming
components and application program interfaces (APIs). TLDT is for web (ASP.NET) and
regular Window applications. MSSQL server is an application used to create computer
databases for the Microsoft windows family of server operating systems. Its server provides
an environment used to generate database that can be accessed from workstations, internet

88
and so on. The server also uses relational database management system that offers a variety
of administrative tools to ease the burdens of database development, maintenance and
administrations [7-9]. The architecture of the operational network for the proposed system
can be seen as a straightforward three-layer comprising Data, Middle and Presentation tiers
as shown in Figure 5. The data tier is basically the server that stores the system‘s
application‘s data and all the query objects used in interacting with the database. Activities at
this level include data retrieval, modification and deletion. [16] The business tier, also known
as middle tier, is the layer that the presentation and the data tiers use for communication.
Typical middle tier components include business logic (such as business rules and data
validation) and data access components and logic including Table and Data Adapters and
Data Readers (ADO. NET), object representations of data (such as LINQ to SQL entity
classes) and common application services (such as authentication, authorization and
personalization). The presentation Layer is the tier at which users interact with the system
and it contains additional applications including data binding components (such as the
binding source and navigator, object representations of data (such as LINQ to SQL entity
classes for use in the presentation tier) and local databases (such as a local database cache).

SEPIA Server: The SEPIA server is a cloud-based server owned by the bank and stores the
user‘s SEPIA service profiles. The server incorporates a callable API server to communicate
with the user application and the ATM terminal. In our case, we have considered Restful
APIs [12] over HTTPS and client-side certificate verification for all communication.

Point-of-Service Terminal: The ATM point-of-service terminal has a unique location

identifier, Loc ID, which is approved and assigned by the bank. The ATM incorporates
network connectivity and can communicate with the bank over secure connection.

User: The user owns a credit/debit card along with a valid Finger & PIN code for
authentication at the point-of-service terminal. The user owns a personal wearable device,
such as the Google Glass, for using the SEPIA service for secure obfuscated PIN
authentication. The user may also choose to use a mobile device for using the SEPIA service.

89
8.4 THE SEPIA PROTOCOL
The SEPIA protocol involves mutual interaction between all three pairs of entities: the user
and the ATM, the ATM and the bank, and the user and the bank, as shown in Figure 1. The
sequence of interactions and messages in the SEPIA protocol is described as follows.

Step 1 [Initiation]: The user, along with the personal mobile or wearable device, approaches
the ATM to perform a secure transaction. The ATM screen displays a ―Touch to begin‖
information screen by default. The user touches the screen (or presses the button) to initiate
the protocol.

Step 2 [ATM Transaction Request]: At this point, the ATM sends an ATM TRAN REQ
message to the bank‘s secure server by the valid fingerimage. The structure of the message is
defined as:

ATM TRAN REQ ⇒ [Req ID, Loc ID] (1)

Here, the Req ID is a request identifier which is generated by the ATM for this current
transaction request. The Loc ID is the unique and verified identifier for the particular ATM
point-of-service assigned by the bank.

Step 3 [Finger Template Generation]: Upon receiving the ATM TRAN REQ message
from the ATM, the bank generates a transaction identifier, Tran ID, for this particular ATM
transaction request. The bank then generates an obfuscated numeric template, Fingerprint
Template, for the transaction to be made at the ATM point-of-service. The Fingerprint
template is an; N-digit numeric pattern, where N ≥ 2P, and P is the length of the Fingerprint
binary code required by the bank for the users. The Fingerprint template is generated using a
random N-digit generator, with a total of P number of digits marked as ‗*‘ at random places.
For example, 8-digit PIN templates for a 4-digit Fingerprint may look like [4 8 * * 2 9 * *],
[* * 4 0 2 * * 6], etc. Finally, the bank creates a record, REC, for the received ATM TRAN
REQ message, and stores it on the local database.

REC ⇒ [Req ID, Loc ID, T ran ID, V alidity, Finger Template, T S, IsUsed] (2)

90
Step 4 [ATM Transaction Response]: Next, the bank server responds to the transaction
request made by the ATM using an ATM TRAN RES message. The structure of the message
is defined as:

ATM TRAN RES ⇒ [T ran ID, V alidity, Fingerprint T emplate] (3)

Here, the Tran ID is the identifier generated by the bank for this particular transaction
request. The Fingerprint Template is the numeric N-digit template generated by the bank.
The bank also sends the Validity token, a timer for the maximum allowed time limit for the
particular Fingerprint template and transaction request for the current user.

Step 5 [QR Code Generation]: Once the ATM receives the ATM TRAN RES message, it
extracts the Tran ID, and generates a quick response (QR) code [15]. The QR code is

generated from the following context: QR Code ⇒ [Loc ID, Req ID, T ran ID] (4)

Here, the Loc ID, Req ID, and Tran ID are the location, request, and transaction identifiers
respectively. The QR code is then displayed on the ATM screen.

Step 6 [QR Code Scan]: At this point, the user is able to see the QR code displayed on the
ATM screen. The user then uses his personal mobile or wearable device running the SEPIA
application to scan the QR code. The advantage of using a personal wearable device, such as
the Google Glass, is that the display of messages in the next phases are only visible to the
interacting user. Upon a successful QR code scan, the Loc ID, Req ID, and Tran ID are
transferred to the user‘s device from the ATM screen.

Step 7 [User Transaction Request]: Once the user scans the QR code on the ATM screen, a
USR TRAN REQ message is created and sent to the bank server over secure communication
channel. The structure of the USR TRAN REQ message is as follows:

USR TRAN REQ ⇒ [Username, P assword, Loc ID, Req ID, T ran ID] (5)

In this message, the Loc ID, Req ID, and Tran ID had been obtained from the QR scan, and
the username and password are the user‘s personal SEPIA service settings which have been
previously saved on the bank‘s website.

91
Step 8 [Transaction Request Verification]: The bank‘s cloud-based server receives the
USR TRAN REQ message from the user‘s personal mobile or wearable device. The bank
then executes the transaction request verification algorithm in the cloud, and responds to the
user‘s personal device. The transaction verification algorithm as mentioned in Table I.

Step 9 [User Transaction Response]: Given that the transaction request verification
algorithm returns a success, the bank server then constructs a USR TRAN RES and sends it
back to the user. The structure of the message is shown as below:

USR TRAN RES ⇒ [Status, [Fingerprint Template, Rem Validity] || [Reason] ] (6)

Figure 8.4 The SELPA Protocol

Here, the status corresponds to the success of the USR TRAN REQ sent earlier. The PIN
Template is obtained from the corresponding REC found in the request verification phase.
Finally, the Rem Validity is the remaining time for the validity of the ATM transaction for
the current user. This is calculated as follows: Rem Validity = REC. Validity - (Current
System Time - TS). Alternatively, if the status is a failure in the request verification phase,
the message includes a Reason for the failure.

Step 10 [Fingerprint Scan]: The fingerprint Scan provides the last level of authentication
for the user. Users only need to place their finger on the scanner for the fingerprint
information to be captured. Once captured, the information is encrypted and transmitted to
the ATM server. The ATM server matches the fingerprint information with the one stored on

92
the database (the template). If a match is confirmed, the server establishes a connection with
the customers‘ bank server and subsequently opens transaction interaction with the customer
via the ATM display screen. On the ATM display screen, the customer can select and
perform any transactions of their choice.

Step 11 [Finger Verification]: The ATM receives the user‘s obfuscated Finger input on the
screen. The Finger Template which the ATM received earlier in the ATM TRAN RES
message is then used to extract the P-digit PIN code obfuscated within the N-digit Finger
Template. The extracted Finger is then used by the ATM to authenticate the user and
completes the SEPIA protocol.

The authentication algorithm for the proposed system follows a simple process as explained
below;

1. User inserts the Finger on the Finger slot. The Finger scanner scan the Finger information
and transmits the encrypted Finger information to the ATM server.

2. The ATM server decrypts the card information to get user‘s account detail; and
subsequently prompts the user through the ATM display screen to supply their PIN.

3. The user keys in their PIN using the keypad, the PIN is encrypted and transmitted to the
ATM server.

4. The ATM server decrypts the PIN and checks with the PIN database for the correctness of
the PIN; and if correct prompts the user to supply their fingerprint information through the
display screen or return ―invalid PIN‖ if not correct and subsequently requests user to retype
their PIN.

5. User places their finger on the fingerprint sensor to take a scan. The fingerprint reader
processes the fingerprint information, encrypts it and transmits it to the ATM server.

6. The ATM server checks with the fingerprint database for correctness of the information;
and if correct establishes a connection with the User‘s bank for transaction operations or
returns ―invalid fingerprint‖ and subsequently takes the user to algorithm number 3.

93
7. When the first transaction is completed, user only needs to supply their fingerprint
information to perform another transaction so long as the card has not been ejected.

8.5 Biometric Enrollment

During Finger registration also called biometric enrolment, a new user supplies their
biometric information to the biometric system. The biometric sensor captures and sends the
information to the ATM client-side biometric processor. [11] The client-side biometric
processor processes the information, and with the help of the cryptographic module, encrypts
and transmits the encrypted information over the network to the ATM server-side processor.
The server-side processor decrypts and processes the encrypted information and extracts
some unique features such as fingerprint minutiae using a software algorithm called feature
extractor. Other identifiers (name and identification number) are added and sent to the
biometric database for storage as a template. This completes the biometric enrollment. In this
work, we have proposed four-finger enrollment, meaning a new user will have to supply
fingerprint information for their two thumbs and two index fingers. This limits the
probability of a denial of service due to system errors or mild fingerprint changes. During
authentication, when the system returns a mismatch for the first finger, users can choose to
try any of the other three fingers.

Figure 8.5 Biometric Enrollment

94
During biometric authentication referred to as user third-factor authentication, a user presents
new biometric sample information to the biometric system through the sensor. The client
processor processes the biometric information and with the cryptographic module encrypts
the information and sends it to the server side processor. At the server side processor, the
supplied information is decrypted and processed. The unique features together with the name
and identification number are extracted and placed on the sample memory map. The server
side processor then queries the biometric database with the sample name and identification
number. The requested templates are supplied and placed on the template map. The processor
now uses a biometric matcher to compare the sample and all four templates associated with
the user for similarities. The matcher returns a match score representing the degree of
similarity between the closest template and the sample.

8.6 Finger Scan Technology

There are five stages involved in finger-scan verification and identification. Fingerprint (FP)
image acquisition, image processing, and location of distinctive characteristics, template
creation and template matching [3]. A scanner takes a mathematical snapshot of a user's
unique biological traits. [9] This snapshot is saved in a fingerprint database as a minutiae file.
The first challenge facing a finger-scanning system is to acquire high-quality image of a
fingerprint. The standard for forensic-quality fingerprinting is images of 500 dots per inch
(DPI). Image acquisition can be a major challenge for finger-scan developers, since the
quality of print differs from person to person and from finger to finger. Some populations are
more likely than others to have faint or difficult-to-acquire fingerprints, whether due to wear
or tear or physiological traits. Taking an image in the cold weather also can have an affect.
Oils in the finger help produce a better print. In cold weather, these oils naturally dry up.
Pressing harder on the platen (the surface on which the finger is placed, also known as a
scanner) can help in this case. Image processing is the process of converting the finger image
into a usable format. This results in a series of thick black ridges (the raised part of the
fingerprint) contrasted to white valleys. At this stage, image features are detected and
enhanced for verification against the stored minutia file. Image enhancement is used to
reduce any distortion of the fingerprint caused by dirt, cuts, scars, sweat and dry skin [3]. The

95
next stage in the fingerprint process is to locate distinctive characteristics. There is a good
deal of information on the average fingerprint and this information tends to remain stable
throughout one‘s life. Fingerprint ridges and valleys form distinctive patterns, such as swirls,
loops, and arches. Most fingerprints have a core, a central point around which swirls, loops,
or arches are curved. These ridges and valleys are characterized by irregularities known as
minutiae, the distinctive feature upon which finger-scanning technologies are based. Many
types of minutiae exits, a common one being ridge endings and bifurcation, which is the
point at which one ridge divides into two. A typical finger-scan may produce between 15 and
20 minutiae. [10] A template is then created. This is accomplished by mapping minutiae and
filtering out distortions and false minutiae. For example, anomalies caused by scars, sweat, or
dirt can appear as minutiae. False minutiae must be filtered out before a template is created
and is supported differently with vendor specific proprietary algorithms. The tricky part is
comparing an enrollment template to a verification template. Positions of a minutia point
may change by a few pixels, some minutiae will differ from the enrollment template, and
false minutiae may be seen as real. Many finger-scan systems use a smaller portion of the
scanned image for matching purposes. One benefit of reducing the comparison area is that
there is less chance of false minutiae information, which would confuse the matching process
and create errors.

Figure 8.6 Finger Scan technology for ATM

96
8.7 Power supply for ATM
The power supply section is required to convert AC signal to DC signal and also to reduce
the amplitude of the signal. The available voltage signal from the mains is 230V/50Hz which
is an AC voltage, but the required is DC voltage (no frequency) with the amplitude of +5V
and +12V for various applications.

Figure 8.7 Power Supply for ATM

The components used in the power supply unit are: step down transformer, bridge rectifier,
capacitor filter, voltage regulator (IC 7805), 330 Ω resistor and LED. Bridge rectifier is
available in IC form (IC DB107). In the present project IC bridge rectifier is used. This
device is ideal for use with printed circuit boards. Electronic filters are electronic circuits,
which perform signal-processing functions, specifically to remove unwanted frequency
components from the signal, to enhance wanted ones. Here a 1000 F capacitor filter is used.

8.8 Summary of This Chapter

The process of converting a varying voltage to a constant regulated voltage is called as
regulation. [10] For the process of regulation we use voltage regulators. A voltage regulator
with only three terminals appears to be a simple device, but it is in fact a very complex
integrated circuit. It converts a varying input voltage into a constant „regulated‟ output
voltage. Voltage Regulators are available in a variety of outputs like 5V, 6V, 9V, 12V and
15V. IC LM 7805 regulator is used.

97
Conclusion
Biometric-based authentication offers several advantages over other authentication methods
such as passwords, passphrase and PINs. This is so because, the fraudster may match
everything but may never match the biometric peculiarities. Biometric tokens are the safest
means of preventing ATM fraud. By further integrating biometric authentication in the ATM
system as a third-factor authentication, we are sure that attackers, impostors and fraudsters as
the case may be, would have a difficult time breaking into peoples‘ accounts. Though there
exists a probability of a possible compromise of the system, the attacker would have to weigh
the attack-resources needed to achieve this with the possible gain; and because our proposed
system offers extremely high attack resources to gain ratio, such efforts may well be an
exercise in futility. The massive adoption and implementation of the system proposed here
will go a long way in solving our ATM security needs. The prototype of ATM systems
authentication based on fingerprint identification will be implemented. Here we will build a
system that will be stable and safe for ATM access. In the results, it will be deduced that the
use of biometric security systems offers a much better authentication of ATM systems and
will take advantages of the stability and reliability of fingerprint characteristics, and a new
biological technology based on the image enhancement algorithm. Additionally, the system
will also include the original verifying methods which are inputting owner's password. These
days, still majority of ATM machine in many countries there are using magnetic card reader,
so there is a need to change a method of authentication in future in order to eliminate the
drawbacks identified in this project. The whole system will br built on the technology of
embedded system which makes the system more safe, reliable and easy to use. ATM
authentication using Fingerprint-based entry is highly susceptible to shoulder-surfing or
observation attacks. Credit/Debit cards are also not resilient to relay and other skimming and
cloning attacks. In this paper, we propose the Secure-Finger Authentication-as-a-Service
(SEPIA), a cloud-based obfuscated Fingerprint-based authentication service for ATMs or
point-of service terminals using personal mobile or wearable devices. We have focused the
security design for SEPIA based on visual privacy of users for a one-time-use Finger
template and address the security vulnerabilities in Fingerprint-based authentication.

98
Appendix
Here we use visual basic software for doing this code .

Admin (E:\atm system)

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class AdminForm
{
public AdminForm()
{
InitializeComponent();
}

#region Default Instance

private static AdminForm defaultInstance;

public static AdminForm Default

{
get
{
if (defaultInstance == null)
{
defaultInstance = new AdminForm();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)

99
 {
defaultInstance = null;
}

#endregion
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
DataTable dt = new DataTable();
string sql;
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
public void GroupBox1_Enter(System.Object sender,
System.EventArgs e)
{

public void Label2_Click(System.Object sender,

System.EventArgs e)
{

public void AdminForm_Load(System.Object sender,

System.EventArgs e)
{
con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
Label11.Text = DateTime.Now.ToString();

txtfname.Enabled = false;
txtlname.Enabled = false;
btnblock.Enabled = false;
btnunblock.Enabled = false;
GroupBox2.Enabled = false;
btnok.Enabled = false;
}

public void Button4_Click(System.Object sender,

System.EventArgs e)
{
txtfname.Enabled = true;
txtlname.Enabled = true;
btnok.Enabled = true;

100
 public void DataGridView1_CellContentClick(System.Object
sender, System.Windows.Forms.DataGridViewCellEventArgs e)
{
int i = default(int);
int j = default(int);

i = e.RowIndex;
j = e.ColumnIndex;
if (j == 0)
{
txtAcctNo.Text =
Convert.ToString((DataGridView1.Rows[i].Cells[j].Value));
txtfnme.Text =
Convert.ToString((DataGridView1.Rows[i].Cells[j + 1].Value));
lblhide.Text = Convert.ToString((DataGridView1.Rows[i].Cells[j
+ 1].Value));
txtlnme.Text = Convert.ToString((DataGridView1.Rows[i].Cells[j
+ 2].Value));
//txtlname.Text = DataGridView1.Rows(i).Cells(j +
2).Value
txtaddr.Text = Convert.ToString((DataGridView1.Rows[i].Cells[j
+ 3].Value));
txtcontact.Text =
Convert.ToString((DataGridView1.Rows[i].Cells[j + 4].Value));
cbGender.Text =
Convert.ToString((DataGridView1.Rows[i].Cells[j + 5].Value));
txtbday.Text = Convert.ToString((DataGridView1.Rows[i].Cells[j
+ 6].Value));
txtPincode.Text =
Convert.ToString((DataGridView1.Rows[i].Cells[j + 7].Value));

}
}

public void Button5_Click(System.Object sender,

System.EventArgs e)
{
con.Open();

System.Data.OleDb.OleDbDataAdapter ad = new
System.Data.OleDb.OleDbDataAdapter("select * from tblinfo", con);
DataSet data = new DataSet();
ad.Fill(data, "info");
DataGridView1.DataSource =
data.Tables["info"].DefaultView;
data.Dispose();

101
 ad.Dispose();
con.Close();
}

public void btnupdate_Click(System.Object sender,

System.EventArgs e)
{
con.Open();

//Dim ad As New OleDb.OleDbDataAdapter("select * from

book", con)

sql = "UPDATE tblinfo SET account_no=\'" +

txtAcctNo.Text + "\',pin_code=\'" + txtPincode.Text + "\',Firstname=\'" +
txtfnme.Text + "\',Lastname=\'" + txtlnme.Text + "\',Address=\'" +
txtaddr.Text + "\',Contact_no=\'" + txtcontact.Text + "\',Gender=\'" +
cbGender.Text + "\'where Firstname=\'" + lblhide.Text + "\'";
cmd.CommandText = sql;
cmd.Connection = con;
cmd.ExecuteNonQuery();
cmd.Dispose();
MessageBox.Show("success");

con.Close();
//Button5_Click(sender, e)
DataGridView1.Refresh();

public void btnok_Click(System.Object sender, System.EventArgs

e)
{
sql = "SELECT * FROM tblinfo where Firstname=\'" +
txtfname.Text + "\'" + "and Lastname= \'" + txtlname.Text + "\'";
try
{
con.Open();
da = new System.Data.OleDb.OleDbDataAdapter(sql,
con);
da.Fill(dt);
DataGridView1.DataSource = dt;

}
catch (Exception ex)
{
Interaction.MsgBox(ex.Message,
MsgBoxStyle.Information, null);
}

con.Close();

102
 btnok.Enabled = false;

public void btnedit_Click(System.Object sender,

System.EventArgs e)
{
GroupBox2.Enabled = true;
btnedit.Enabled = false;
Button4.Enabled = false;
btnblock.Enabled = true;
btnunblock.Enabled = true;

public void btncancel_Click(System.Object sender,

System.EventArgs e)
{
GroupBox2.Enabled = false;
btnedit.Enabled = true;
Button4.Enabled = true;
}

public void Button1_Click(System.Object sender,

System.EventArgs e)
{
Login_frm.Default.Show();
this.Hide();

public void btnblock_Click(System.Object sender,

System.EventArgs e)
{
con.Open();
System.Data.OleDb.OleDbDataAdapter ad = new
System.Data.OleDb.OleDbDataAdapter("select * from tblinfo", con);

sql = "UPDATE tblinfo SET type=\'" + "Block" + "\'" + "

Where Firstname =\'" + txtfnme.Text + "\'";
cmd.CommandText = sql;
cmd.Connection = con;
cmd.ExecuteNonQuery();
cmd.Dispose();
MessageBox.Show("success");

con.Close();
Button5_Click(sender, e);
}

103
 public void btnunblock_Click(System.Object sender,
System.EventArgs e)
{
con.Open();
System.Data.OleDb.OleDbDataAdapter ad = new
System.Data.OleDb.OleDbDataAdapter("select * from tblinfo", con);
sql = "UPDATE tblinfo SET type=\'" + "Active" + "\'" + "
Where Firstname =\'" + txtfnme.Text + "\'";
cmd.CommandText = sql;
cmd.Connection = con;
cmd.ExecuteNonQuery();
cmd.Dispose();
MessageBox.Show("success");
con.Close();
Button5_Click(sender, e);
}
}
}

Balance inquire
using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Balanceinq
{
public Balanceinq()
{
InitializeComponent();
}
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
DataSet ds = new DataSet();

104
 System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
string sql;

public void Balanceinq_Load(System.Object sender,

System.EventArgs e)
{
Label2.Text = DateTime.Now.ToString();
}

public void Button1_Click(System.Object sender,

System.EventArgs e)
{
string sql = default(string);
DataTable Log_in = new DataTable();
try
{
if (txtpin.Text == "")
{
MessageBox.Show("Pls Enter Your Pin");
}
else
{
con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
sql = "SELECT * FROM tblinfo where pin_code
= " + txtpin.Text + "";

cmd.Connection = con;
cmd.CommandText = sql;
da.SelectCommand = cmd;
da.Fill(Log_in);
if (Log_in.Rows.Count > 0)
{
string balance = default(string);

balance = (string)
(Log_in.Rows[0]["balance"]);

Receipt.Default.Show();
//Receipt.lblaccno.Text =
lblaccno.Text
Receipt.Default.lblbal.Text = balance;
Receipt.Default.Label4.Hide();
Receipt.Default.Label3.Hide();
Receipt.Default.lbldep.Hide();
Receipt.Default.lblwith.Hide();
Receipt.Default.Label6.Hide();
Receipt.Default.lblnewbal.Hide();

105
 this.Hide();
}
else
{
MessageBox.Show("Pincode is
incorrect");

}
}

}
catch (Exception ex)
{

MessageBox.Show(ex.Message);
}

txtpin.Text = "";
}
}
}

Deposit

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Deposit
{
public Deposit()
{
InitializeComponent();

}’

106
#region Default Instance

private static Deposit defaultInstance;

public static Deposit Default

{
get
{
if (defaultInstance == null)
{
defaultInstance = new Deposit();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)
{
defaultInstance = null;
}

#endregion
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
DataSet dset = new DataSet();
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
string balance;
int num1;
int num2;
int total;

public void Deposit_Load(System.Object sender,

System.EventArgs e)
{
lbldate.Text = DateTime.Now.ToString();
lblaccno.Text = Mainmenu.Default.lblaccno.Text;
}

public void btnok_Click(System.Object sender, System.EventArgs

e)
{

107
 string sql = default(string);
DataTable Log_in = new DataTable();

try
{
con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
sql = "SELECT * FROM tblinfo where account_no = " +
lblaccno.Text + "";

cmd.Connection = con;
cmd.CommandText = sql;
da.SelectCommand = cmd;
da.Fill(Log_in);
if (Log_in.Rows.Count > 0)
{
balance = (string)
(Log_in.Rows[0]["balance"]);
num1 = int.Parse(balance);
num2 = int.Parse(txtamount.Text);

if (num2 > 25000)

{
MessageBox.Show("You can Only Deposit Php
25,000!");
}
else if (num2 < 200)
{
MessageBox.Show(" Mininum Deposit is
200");
}
else
{
total = num1 + num2;

Receipt.Default.Show();
Receipt.Default.lblbal.Text = balance;
Receipt.Default.Label3.Hide();
Receipt.Default.lblwith.Hide();
Receipt.Default.lbldep.Text =
num2.ToString();
Receipt.Default.lblnewbal.Text =
total.ToString();
Receipt.Default.Label5.Show();
Receipt.Default.Label6.Show();

Receipt.Default.lblbal.Show();
Receipt.Default.Label3.Hide();
Receipt.Default.lblwith.Hide();

108
 Receipt.Default.lbldep.Show();
Receipt.Default.lblnewbal.Show();
//MsgBox("success")
Receipt.Default.lblname.Text =
Mainmenu.Default.lblname.Text;
this.Hide();

}
}
else
{

}
catch (Exception)
{
MessageBox.Show(" Pls. Enter Ammount!");
//MsgBox(ex.Message)
}
txtamount.Text = "";

public void LinkLabel1_LinkClicked(System.Object sender,

System.Windows.Forms.LinkLabelLinkClickedEventArgs e)
{
Mainmenu.Default.Show();
this.Hide();

public void txtamount_TextChanged(System.Object sender,

System.EventArgs e)
{

public void lbldate_Click(System.Object sender,

System.EventArgs e)
{

}
}

109
Login Form

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Login_frm
{
public Login_frm()
{
InitializeComponent();

#region Default Instance

private static Login_frm defaultInstance;

public static Login_frm Default

{
get
{
if (defaultInstance == null)
{
defaultInstance = new Login_frm();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)
{
defaultInstance = null;
}

110
#endregion
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
DataSet ds = new DataSet();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
public void btnlogin_Click(System.Object sender,
System.EventArgs e)
{
string sql = default(string);
DataTable Log_in = new DataTable();
try
{
if (txtpin.Text == "")
{
MessageBox.Show("Pls Enter both Fields");

}
else
{
con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
sql = "SELECT * FROM tblinfo where pin_code
= " + txtpin.Text + "";

cmd.Connection = con;
cmd.CommandText = sql;
da.SelectCommand = cmd;
da.Fill(Log_in);
if (Log_in.Rows.Count > 0)
{
string Type;
string Fullname = default(string);
string accno = default(string);
Type = (string)
(Log_in.Rows[0]["type"]);
Fullname = (string)
(Log_in.Rows[0]["Firstname"]);
accno =
Convert.ToString((Log_in.Rows[0]["account_no"]));
if (Type == "admin")
{
MessageBox.Show("Welcome " +
Fullname + " you login as Administrator ");
AdminForm.Default.Show();
this.Hide();

111
 }
else if (Type == "Block")
{
MessageBox.Show("Your account is
currently Block");
MessageBox.Show("Contact the
Administrator for Help");

}
else
{
MessageBox.Show("Welcome " +
Fullname);

Mainmenu.Default.lblname.Text =
Fullname;
Mainmenu.Default.lblaccno.Text =
accno;
Mainmenu.Default.Show();
this.Hide();

}
else
{
MessageBox.Show("Yuo are Not
Registered!!!");
MessageBox.Show("Pls Register if You
are New!");

}
}
}
catch (Exception ex)
{
MessageBox.Show(ex.Message);
}

txtpin.Text = "";

public void llblreg_LinkClicked(System.Object sender,

System.Windows.Forms.LinkLabelLinkClickedEventArgs e)
{
Regs_frm.Default.Show();

112
 }

public void Login_frm_Load(System.Object sender,

System.EventArgs e)
{

public void GroupBox1_Enter(System.Object sender,

System.EventArgs e)
{

}
}

Login Menu

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Mainmenu
{
public Mainmenu()
{
InitializeComponent();

#region Default Instance

private static Mainmenu defaultInstance;

113
 public static Mainmenu Default
{
get
{
if (defaultInstance == null)
{
defaultInstance = new Mainmenu();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)
{
defaultInstance = null;
}

#endregion
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
DataSet ds = new DataSet();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
string sql;
public void Label4_Click(System.Object sender,
System.EventArgs e)
{

public void Button4_Click(System.Object sender,

System.EventArgs e)
{
Login_frm.Default.Show();
this.Hide();
}

public void Mainmenu_Load(System.Object sender,

System.EventArgs e)
{
lbldate.Text = DateTime.Now.ToString();

114
 public void Button1_Click(System.Object sender,
System.EventArgs e)
{

string sql = default(string);

DataTable Log_in = new DataTable();
try
{

con.ConnectionString = "Provider=Microsoft.ACE.OLEDB.12.0;
Data Source=" +
Application.StartupPath + "\\ATMsystem.accdb";
sql = "SELECT * FROM tblinfo where account_no = " +
lblaccno.Text + "";

cmd.Connection = con;
cmd.CommandText = sql;
da.SelectCommand = cmd;
da.Fill(Log_in);
if (Log_in.Rows.Count > 0)
{
string balance = default(string);

balance = (string)
(Log_in.Rows[0]["balance"]);

Receipt.Default.Show();
Receipt.Default.lblname.Text = lblname.Text;
//Receipt.lblaccno.Text = lblaccno.Text
Receipt.Default.lblbal.Text = balance;
Receipt.Default.Label4.Hide();
Receipt.Default.Label3.Hide();
Receipt.Default.lbldep.Hide();
Receipt.Default.lblwith.Hide();
Receipt.Default.Label6.Hide();
Receipt.Default.lblnewbal.Hide();

this.Hide();

}
else
{
MessageBox.Show("Pincode is incorrect");

115
 catch (Exception ex)
{

MessageBox.Show(ex.Message);
}

public void btnwith_Click(System.Object sender,

System.EventArgs e)
{
Withdraw.Default.Show();
this.Hide();

public void btndep_Click(System.Object sender,

System.EventArgs e)
{
Deposit.Default.Show();
this.Hide();
}

public void lblname_Click(System.Object sender,

System.EventArgs e)
{

public void GroupBox1_Enter(System.Object sender,

System.EventArgs e)
{

public void Button2_Click(System.Object sender,

System.EventArgs e)
{

}
}
}

116
Receipt
using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Receipt
{
public Receipt()
{
InitializeComponent();
}

#region Default Instance

private static Receipt defaultInstance;

public static Receipt Default
{
get
{
if (defaultInstance == null)
{
defaultInstance = new Receipt();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)
{
defaultInstance = null;
}

#endregion

117
 System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
DataTable dt = new DataTable();
string sql;
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();

public void Receipt_Load(System.Object sender,

System.EventArgs e)
{

con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
lbldate.Text = DateTime.Now.ToString();

public void Button1_Click(System.Object sender,

System.EventArgs e)
{
if (lblnewbal.Text == "")
{

}
else
{

con.Open();

int total = int.Parse(lblnewbal.Text);

System.Data.OleDb.OleDbDataAdapter ad = new
System.Data.OleDb.OleDbDataAdapter("select * from tblinfo", con);

sql = "UPDATE tblinfo SET balance=\'" +

total.ToString() + "\'" + " Where Firstname=\'" + lblname.Text + "\'";
cmd.CommandText = sql;
cmd.Connection = con;
cmd.ExecuteNonQuery();
cmd.Dispose();

con.Close();

118
 if (MessageBox.Show("Do You Want to Continue Your
transaction??", "Continue", MessageBoxButtons.YesNo,
MessageBoxIcon.Question) == System.Windows.Forms.DialogResult.Yes)
{
Mainmenu.Default.Show();
}
else
{
MessageBox.Show("Thank You Come Again");

Login_frm.Default.Show();

}
this.Close();
}
}
}

Registration Form

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Regs_frm
{
public Regs_frm()
{

constr = "Provider=Microsoft.ACE.OLEDB.12.0;Data
Source=" + Application.StartupPath + "\\ATMsystem.accdb";
conn = new System.Data.OleDb.OleDbConnection(constr);

InitializeComponent();
}

#region Default Instance

private static Regs_frm defaultInstance;

119
 public static Regs_frm Default
{
get
{
if (defaultInstance == null)
{
defaultInstance = new Regs_frm();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

static void defaultInstance_FormClosed(object sender,

FormClosedEventArgs e)
{
defaultInstance = null;
}

#endregion
string constr;
System.Data.OleDb.OleDbDataAdapter adapt;
System.Data.OleDb.OleDbDataAdapter adapt1 = new
System.Data.OleDb.OleDbDataAdapter();
System.Data.OleDb.OleDbConnection conn;
DataSet dset = new DataSet();

public void btnRegister_Click(System.Object sender,

System.EventArgs e)
{
if (txtAcctNo.Text == "" && txtPincode.Text == "" &&
txtcontact.Text == "" && txtfname.Text == "" && txtlname.Text == "" &&
txtaddr.Text == "" && txtcontact.Text == "" && cbGender.Text == "" &&
cbday.Text == "" && cbmonth.Text == "" && cbyear.Text == "")
{
MessageBox.Show("Enter All Fields");

}
else if (txtAcctNo.Text == "" || txtPincode.Text == ""
|| txtcontact.Text == "" || txtfname.Text == "" || txtlname.Text == "" ||
txtaddr.Text == "" || txtcontact.Text == "" || cbGender.Text == "" ||
cbday.Text == "" || cbmonth.Text == "" || cbyear.Text == "")
{
MessageBox.Show("Pls Complete all Fields");

120
 }
else
{
System.Data.OleDb.OleDbDataAdapter adapt1 = new
System.Data.OleDb.OleDbDataAdapter("select * from tblinfo where
Firstname=\'" + txtfname.Text + "\'", conn);
DataSet dset1 = new DataSet();
adapt1.Fill(dset1);
if (dset1.Tables[0].Rows.Count != 0)
{
MessageBox.Show("Account name already
exist");
}
else
{
string dbcommand = "INSERT into tblinfo
(account_no, Firstname, Lastname, Address, Contact_no, Gender, Birthday,
pin_code , type, balance)" + " VALUES (\'" + (txtAcctNo.Text + "\',\'") +
txtfname.Text + "\',\'" + txtlname.Text + "\',\'" + txtaddr.Text + "\',\'"
+ txtcontact.Text + "\',\'" + cbGender.Text + "\',\'" + (cbmonth.Text +
cbday.Text + cbyear.Text) + "\',\'" + txtPincode.Text + "\',\'" + "Active"
+ "\',\'" + "1000" + "\')";
System.Data.OleDb.OleDbDataAdapter adapt =
new System.Data.OleDb.OleDbDataAdapter(dbcommand, conn);
DataSet dset = new DataSet();
adapt.Fill(dset);
MessageBox.Show("You Have Succesfully
Registered!");
this.Hide();
Login_frm.Default.Show();
}
}
}

public void Regs_frm_Load(System.Object sender,

System.EventArgs e)
{
lbldate.Text = DateTime.Now.ToString();

conn.ConnectionString = constr;
conn.Open();
}

public void Button1_Click(System.Object sender,

System.EventArgs e)
{
this.Close();
Login_frm.Default.Show();
}
}

121
Withdraw

using System.Diagnostics;
using System;
using System.Xml.Linq;
using System.Windows.Forms;
using System.Collections;
using System.Drawing;
using Microsoft.VisualBasic;
using System.Data;
using System.Collections.Generic;
using System.Linq;

namespace atmsystem
{
public partial class Withdraw
{
public Withdraw()
{
InitializeComponent();

#region Default Instance

private static Withdraw defaultInstance;

/// <summary>
/// Added by the VB.Net to C# Converter to support default
instance behavour in C#
/// </summary>
public static Withdraw Default
{
get
{
if (defaultInstance == null)
{
defaultInstance = new Withdraw();
defaultInstance.FormClosed += new
FormClosedEventHandler(defaultInstance_FormClosed);
}

return defaultInstance;
}
}

122
 static void defaultInstance_FormClosed(object sender,
FormClosedEventArgs e)
{
defaultInstance = null;
}

#endregion
System.Data.OleDb.OleDbDataAdapter da = new
System.Data.OleDb.OleDbDataAdapter();
System.Data.OleDb.OleDbConnection con = new
System.Data.OleDb.OleDbConnection();
DataSet dset = new DataSet();
System.Data.OleDb.OleDbCommand cmd = new
System.Data.OleDb.OleDbCommand();
string balance;
int num1;
int num2;
int total;

public void Withdraw_Load(System.Object sender,

System.EventArgs e)
{
lbldate.Text = DateTime.Now.ToString();
lblaccno.Text = Mainmenu.Default.lblaccno.Text;
}

public void btnok_Click(System.Object sender, System.EventArgs

e)
{

string sql = default(string);

DataTable Log_in = new DataTable();

try
{

con.ConnectionString =
"Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + Application.StartupPath
+ "\\ATMsystem.accdb";
sql = "SELECT * FROM tblinfo where account_no = "
+ lblaccno.Text + "";

cmd.Connection = con;
cmd.CommandText = sql;
da.SelectCommand = cmd;
da.Fill(Log_in);
if (Log_in.Rows.Count > 0)
{
balance = (string)
(Log_in.Rows[0]["balance"]);

123
 num1 = int.Parse(balance);
num2 = int.Parse(txtamount.Text);

if (num2 > 25000)

{
MessageBox.Show("You can Only Withdraw
Php 25,000");
}
else if (num2 < 200)
{
MessageBox.Show(" Mininum withdrawal
is 200");
}
else if (num1 < num2)
{
MessageBox.Show(" Insuffiecient
balance");

}
else
{
total = num1 - num2;

Receipt.Default.Show();

Receipt.Default.lblbal.Text = balance;
Receipt.Default.Label4.Hide();
Receipt.Default.lbldep.Hide();
Receipt.Default.lblwith.Text =
num2.ToString();
Receipt.Default.lblnewbal.Text =
total.ToString();

Receipt.Default.Label5.Show();
Receipt.Default.Label6.Show();

Receipt.Default.lblbal.Show();
Receipt.Default.Label4.Hide();
Receipt.Default.lbldep.Hide();
Receipt.Default.lblwith.Show();
Receipt.Default.lblnewbal.Show();
//MsgBox("success")
Receipt.Default.lblname.Text =
Mainmenu.Default.lblname.Text;
this.Hide();

}
}

124
 else
{

}
catch (Exception)
{
MessageBox.Show(" Pls. Enter Ammount!");
//MsgBox(ex.Message)
}
txtamount.Text = "";

public void LinkLabel1_LinkClicked(System.Object sender,

System.Windows.Forms.LinkLabelLinkClickedEventArgs e)
{
Mainmenu.Default.Show();
this.Hide();
}
}
}

Fingerprint Matching
using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Text;
using System.Windows.Forms;
using System.Drawing.Imaging;
using System.IO;

namespace Fingerprint_Based_Atm_System
{
public partial class Form1 : Form
{
Bitmap bitmap1, bitmap2;
public Form1()
{
InitializeComponent();

125
 }

private void button1_Click(object sender, EventArgs e)

{
OpenFileDialog openflg = new OpenFileDialog();
if (openflg.ShowDialog() == DialogResult.OK)
{
pictureBox1.ImageLocation = openflg.FileName;
bitmap1 = new Bitmap(openflg.FileName);
}
}
private void button2_Click(object sender, EventArgs e)
{
OpenFileDialog openflg = new OpenFileDialog();
if (openflg.ShowDialog() == DialogResult.OK)
{
pictureBox2.ImageLocation = openflg.FileName;
bitmap2 = new Bitmap(openflg.FileName);
}
}
private void button3_Click(object sender, EventArgs e)
{
bool compare = ImageCompareString(bitmap1, bitmap2);
if (compare == true)
{
MessageBox.Show("Thank You. Your Transation is
ready");
}
else
{
MessageBox.Show("Please Enter The Right Finger");

}
}

private bool ImageCompareString(Bitmap bitmap11, Bitmap

bitmap21)
{
MemoryStream ms = new MemoryStream();
bitmap11.Save(ms, ImageFormat.Png);
String firstbitmap =
Convert.ToBase64String(ms.ToArray());
ms.Position = 0;
bitmap21.Save(ms, ImageFormat.Png);
String secondbitmap =
Convert.ToBase64String(ms.ToArray());

126
 if (firstbitmap.Equals(secondbitmap))
{
return true;
}
else
{
return false;
}
}
}
}

127
Reference
[1] Oh, Choonsuk. "Study On The Center Of Rotation Method Based On Minimum Spanning
Tree Matching Algorithm For Fingerprint Recognition". Optical Engineering 43.4 (2004): 822.
Web.

[2] Perminov, V. N. and A. M. Fartukov. "Method For Fingerprint Minutiae Matching Based On
Their Alignment". Pattern Recognition and Image Analysis 17.4 (2007): 631-638. Web.

[3] Shahan, Timothy A. and Christopher A. Podlesnik. "Divided Attention Performance And The
Matching Law". Learning & Behavior 34.3 (2006): 255-261. Web.

[4] Tuan, Luu Trong. "Matching And Stretching Learners‟ Learning Styles". Journal of
Language Teaching and Research 2.2 (2011): n. pag. Web.

[5] Beresford, John. "Matching Teaching To Learning". Curriculum Journal 10.3 (1999): 321-
344. Web.

[6] "Fingerprint, Palms And Soles; An Introduction To Dermatoglyphics". Postgraduate Medical

Journal 39.448 (1963): 104-105. Web.

[7] Zakaria, M. K. et al. "Introduction Of Fingerprint Recognition Method For Machinery Faults
Diagnosis". AMM 575 (2014): 829-832. Web.

[8] "Fingerprint Based Authentication System Using ARM7". IJSR 5.5 (2016): 440-443. Web.

[9] Willis, A.J. and L. Myers. "A Cost-Effective Fingerprint Recognition System For Use With
Low-Quality Prints And Damaged Fingertips". Pattern Recognition 34.2 (2001): 255-270. Web.

[10] Singh, Dayashankar, Dr. P.K. Singh, and Dr. R.K. Shukla. "Fingerprint Recognition System
Based On Mapping Approach". International Journal of Computer Applications 5.2 (2010): 1-5.
Web.

[11] Miron, Radu Florin and Mihai Hulea. "Access Control Based On A Distributed Fingerprint
Recognition System". Applied Mechanics and Materials 555 (2014): 759-764. Web.

[12] LIU, Bo. "Fingerprint Image Enhancement Using Mixed Filters". Journal of Computer
Applications 28.7 (2008): 1892-1895. Web.

[13] Koichi I., Ayumi M., Takafumi A., Hiroshi N., Koji Kobayashi, and Tatsuo H. (2005) ―A
Fingerprint Recognition Algorithm Combining Phase-Based Image Matching and Feature-
BasedMatching‖.

[14] Anton S. (2002) ―Sorting it out: Machine learning and finger-prints‖, Paper presented at the
seminar on Telematik finger-print, Siemens Corporate Technology, Munich, Germany.

128
[15] Thai R. (2003) ―Fingerprint Image Enhancement and Minutiae Extraction―, Unpublished
B.Sc Thesis, School of Computer Sci-ence and Software Engineering, The University of Western
Aus-tralia, Australia.

[16] ―Characteristics and Application of the Fingerprint Recognition Systems‖, I. Plajh et al,
MEASUREMENT SCIENCE REVIEW, Volume 3, Section 2, 2003.

129