A TECHNICAL SEMINAR REPORT

ON
AN ATM WITH AN EYE
Submitted in partial fulfilment of the requirements for the award of degree of

BACHELOR OF TECHNOLOGY
In
ELECTRONICS AND COMMUNICATION ENGINEERING
By

P. NISHITHA

(19D21A0437)

SRIDEVI WOMEN’S ENGINEERING COLLEGE

Department of Electronics and Communication Engineering
Accredited by NBA & NAAC. An ISO 9001:2015 certified institution

(Affiliated to JNTU, Hyderabad, Approved by AICTE, New Delhi)

V.N. PALLY, GANDIPET , R.R DIST.-500075

1
 SRIDEVI WOMEN’S ENGINEERING COLLEGE
Department of Electronics and Communication Engineering
Accredited by NBA & NAAC. An ISO 9001:2015 certified institution

(Affiliated to JNTU, Hyderabad, Approved by AICTE, New Delhi)

V.N.PALLY, GANDIPET ,R.R DIST.-500075

2022-2023

BONAFIDE CERTIFICATE
This is to certify that the report entitled “ AN ATM WITH AN EYE” that is being submitted
by P . NISHITHA (19D21A0437) in partial fulfilment for the award of bachelor of technology in
Electronics and Communication Engineering to Jawaharlal Nehru Technological University,
Hyderabad is a record of bonafide work carried out by her our guidance and supervision.
The results contents embodied is the report have not been submitted to any other University or
institute for the award of any degree or diploma.

CO-ORDINATOR

H . Shanthi

Assistant professor

2
 DECLARATION
I hereby declare that the work described in this technical report entitled “AN ATM WITH
AN EYE” which is being submitted by me in partial fulfilment for the award of Bachelor of
Technology in the department of ELECTRONICS AND COMMUNICATION
ENGINEERING to the SRIDEVI WOMEN’S ENGINEERING COLLEGE Affiliated to
Jawaharlal Nehru Technological University Hyderabad , Kukatpally.

P. NISHITHA ( 19D21A0437)

3
 ACKNOWLEDGEMENT
With great pleasure I want to take this opportunity to express my gratitude to all the people
who helped in making this technical seminar work a grand success.

I am grateful to our technical seminar coordinator Mrs. H .SHANTHI, for her

valuable suggestions and guidance during the execution of this technical seminar.
I would like to thank Dr. A. NARMADA, Head of the Department,
Department of Electronics and Communications Engineering , for being moral support
throughout the period of my study.

I express my deep sense of gratitude to Dr. B.L. MALLESHWARI, principal, SRIDEVI

WOMEN’S ENGINEERING COLLEGE. For her constant guidance throughout my technical
seminar.

I would like to thank the Teaching and Non -Teaching staff of ECE Department for
sharing their knowledge with us.

Finally , I express my sincere thanks to the Management of Sridevi women’s engineering

college , for providing excellent facilities in the college for their continuous support towards
my achievements.

P NISHITHA
(19D21A0437)

4
 ABSTRACT

Automated Teller Machines also known as ATM’s are widely used nowadays by each and
everyone. The ATM machine (Automated Teller Machine) is an electronic device that is used
by the banks to perform banking tasks like withdrawal of money, transferring of money, and
many to get many information about a user's bank account without the need to visit a bank.
This System revolutionised the way of transactions. There were no long lines of queue in
front of the bank for a simple withdrawal of money. The number of ATM’s a bank has can be
a factor in considering the strength of a bank. As there is increase in the number of ATM’s,
there is also increase in the fraudulent activities in the ATM. The main motivation of this
project is to increase the security feature of the use of ATM. The current method uses static
key (PIN) for security. The proposed method uses Face-id as a key incorporated with current
method. The advantages can be found as that the face-id is unique for everybody; it cannot
be used by anybody other than the user.

5
 CONTENTS

LIST OF CONTENTS PAGE NO

ACKNOWLEDGEMENT 4
ABSTRACT 5
LIST OF FIGURES
CHAPTER 1 INTRODUCTION 7
1.1 Automated Teller Machine (ATM)……………………………….... 7
1.2 How ATM works with Biometrics…………………….………….... 8
1.3 Purpose and Benefits……………….................................................. 8-9
CHAPTER 2 LITERATURE SURVEY

CHAPTER 3 10
2.1 Algorithms used ……………………….…………………………….10
2.2 Eigenface Face Recognizer Algorithm………………………………11

6
7
 CHAPTER 1
INTRODUCTION

Nowadays, the uses of ATMs are inevitable in our day-to-day lives. Before the introduction
of ATM, people used to travel with cash in their hands. This caused the risk of getting robbed
or the cash might get lost due to the negligence of the user. To overcome this issues ATM
played a big role. It is like cash in a card. As the ATM technology got improved the idea of
misusing it also got improved. The main disadvantage of this system is that if the PIN is
known anybody can use it to take money.

To address this issue, Face-id is used as a key. The main use of the Biometric is that it is
unique for a person. The proposed innovation is not an alternative for the conventional ATM
security system. The proposed innovation is considered as an additional support for the
existing method. In India, nearly 2000 ATM crimes are occurring every year. The proposed
method will surely reduce this rate to a greater extent.

1.1 Automated Teller Machine

An automated teller machine (ATM) is an electronic banking outlet that allows

customers to complete basic transactions without the aid of a branch representative or
teller. Anyone with a credit card or debit card can access cash at most ATMs, either in
the USA or abroad.

ATMs are convenient, allowing consumers to perform quick self-service transactions

such as deposits, cash withdrawals, bill payments, and transfers between accounts.
Fees are commonly charged for cash withdrawals by the bank where the account is
located, by the operator of the ATM, or by both. Some or all of these fees can be
avoided by using an ATM operated directly by the bank that holds the account. Using
an ATM abroad can cost more than using one in the USA.

ATMs are known in different parts of the world as automated bank machines (ABMs)
or cash machines.

8
1.2 How ATM works with Biometrics

The ATM prompts the user to insert the card. When the card is inserted the
customer is requested for personal identification number(PIN). The customer
inputs his/her PIN. When the PIN matches the account then an ordinary video
camera captures an image of the customer’s face. This image is converted in to a
digital code, which is compared with one already stored for that individual if the
system perceives a match, the customer can proceed to the with drawl of cash.

When the match is made with PIN but not the images, the bank could limit
transactions in the manner agreed upon by the customer when the account was
opened, and could store the image of the user for later examination by the bank
officials. ATM reads and store customer’s data on a smart card.

1.3 Purpose and Benefits

Face recognition finds its application in a variety of fields such as homeland
security, criminal identification, human-computer interaction, privacy security, etc.
The face recognition feature inhibits access of account through stolen or fake
cards. The card itself is not enough to access account as it requires the person as
well for the transaction to proceed. Eigenface based method is used for the face
recognition. However, the drawback of using eigenface based method is that it can
sometimes be spoofed by the means of fake masks or photos of an account holder.
To overcome this problem 3D face recognition methods can be used. However, its
computation cost is high and requires large storage space which makes it very
difficult to store information about a large number of users and 3D masks can also
be used to spoof the 3D facial recognition based model. 3D printing is mostly used
for such attacks. These drawbacks can be easily overcome by using One-Time
passwords (OTP). OTP ensures that the user is authentic by sending the randomly
generated 6-digit code to the registered mobile number of the corresponding
account holder. In addition, the user will not have to remember PIN. It prevents the
fraudulent attacks like:

9
 Eavesdropping The ATM card or PIN of a user can be spied upon and can
be accessed easily by obtaining the card by faulty means. This can lead to
some serious consequences.

 Spoofing There is a possibility that, when a user enters the PIN during the
transaction process, a hacker fakes as the authorized site and prompts the
user to re-enter PIN due to a system error. When a user complies with the
instruction the hacker stores the data and uses it for his future peccadilloes
intentions. This man-in-the-middle (hacker) attack is futile because new
password is temporarily assigned in every new transaction.

 Brute-force attack Using the brute force, if we try to crack the current
static four digit PIN it can be done in 9999 attempts, thus weakening the
security. In our model a 6-digit code is sent to a registered number, thus
increasing the security and reducing the chances of cracking the code
using brute force.

10
 CHAPTER 2
LITERATURE SURVEY

For most of the past ten years, the majority of ATMs used worldwide ran under IBM’s now-
defunct OS/2. However, IBM hasn’t issued a major update to the operating system in over six
years. Movement in the banking world is now going in two directions: Windows and Linux.
NCR, a leading world-wide ATM manufacturer, recently announced an agreement to use
Windows XP Embedded in its next generation of personalized ATMs Windows XP
Embedded allows OEMs to pick and choose from the thousands of components that make up
Windows XP Professional, including integrated multimedia, networking and database
management functionality. This makes the use of off-the-shelf facial recognition code more
desirable because it could easily be compiled for the Windows XP environment and the
networking and database tools will already be in place.

For less powerful ATMs, KAL, a software development company based in Scotland,
provides Kalignite CE, which is a modification of the Windows CE platform. This allows
developers that target older machines to more easily develop complex user interaction
systems. Many financial institutions are relying on a third choice, Windows NT, because of
its stability and maturity as a platform. On an alternative front, the largest bank in the south
of Brazil, Banrisul, has installed a custom version of Linux in its set of two thousand ATMs,
replacing legacy MS-DOS systems.

The ATMs send database requests to bank servers which do the bulk of transaction
processing This model would also work well for the proposed system if the ATMs
processors were not powerful enough to quickly perform the facial recognition algorithms. In
terms of the improvement of security standards, MasterCard is spearheading an effort to
heighten the encryption used at ATMs. For the past few decades, many machines have used
the Data Encryption Standard developed by IBM in the mid 1970s that uses a 56-bit key.
DES has been shown to be rather easily cracked, however, given proper computing hardware.

In recent years, a “Triple DES” scheme has been put forth that uses three such keys, for an
effective 168-bit key length. MasterCard now requires new or relocated ATMs to use the
Triple DES scheme, and by April, 2005, both Visa and MasterCard will require that any
ATM that supports their cards must use Triple DES. ATM manufacturers are now developing
newer models that support Triple DES natively; such redesigns may make them more
amenable to also including snapshot cameras and facial recognition software, more so than
they would be in regards to retrofitting preexisting machines.

11
There are hundreds of proposed and actual implementations of facial recognition technology
from all manner of vendors for all manner of uses. However, for the model proposed in this
paper, we are interested only in the process of facial verification – matching a live image to a
predefined image to verify a claim of identity – not in the process of facial evaluation –
matching a live image to any image in a database. Further, the environmental conditions
under which the verification takes place – the lighting, the imaging system, the image profile,
and the processing environment – would all be controlled within certain narrow limits,
making hugely robust software unnecessary .

One leading facial recognition algorithm class is called image template based. This method
attempts to capture global features of facial images into facial templates. Neural networks,
among other methods, are often used to construct these templates for later matching use. An
alternative method, called geometry-based, is to explicitly examine the individual features of
a face and the geometrical relationship between those features (Gross.) What must be taken
into account, though, are certain key factors that may change across live images: illumination,
expression, and pose (profile). A study was recently conducted of leading recognition
algorithms, notably one developed by two researchers at MIT, Baback Moghaddam and Alex
Pentland, and one a commercial product from Identix called FaceIt.

The MIT program is based on Principal Feature Analysis, an adaptation of template based
recognition. FaceIt’s approach uses geometry-based local feature analysis. Both algorithms
have to be initialized by providing the locations of the eyes in the database image, from
which they can create an internal representation of the normalized face. It is this
representation to which future live images will be compared. In the study, it was found that
both programs handled changes in illumination well. This is important because ATM use
occurs day and night, with or without artificial illumination. Likewise, the programs allowed
general expression changes while maintaining matching success.

However, extreme expressions, such as a scream profile, or squinted eyes, dropped the
recognition rates significantly. Lastly, matching profile changes worked reasonably well
when the initial training image(s) were frontal, which allowed 70-80% success rates for up to
45 degrees of profile change however, 70-80% success isn’t amenable to keeping ATM users
content with the system.

12
 The natural conclusion to draw, then, is to take a frontal image for the bank database, and to
provide a prompt to the user, verbal or otherwise, to face the camera directly when the ATM
verification process is to begin, so as to avoid the need to account for profile changes. With
this and other accommodations, recognition rates for verification can rise above 90%. Also
worth noting is that FaceIt’s local feature analysis method handled variations in the test cases
slightly better than the PGA system used by the MIT researchers. Another paper shows more
advantages in using local feature analysis systems. For internal representations of faces, LFA
stores them topographically; that is, it maintains feature relationships explicitly. Template
based systems, such as PGA, do not.

The advantages of LFA are that analysis can be done on varying levels of object grouping,
and that analysis methods can be independent of the topography. In other words, a system can
examine just the eyes, or the eyes nose and mouth, or ears, nose, mouth and eyebrows, and so
on, and that as better analysis algorithms are developed, they can fit within the data
framework provided by LFA. The conclusion to be drawn for this project, then, is that facial
verification software is currently up to the task of providing high match rates for use in ATM
transactions. What remains is to find an appropriate open-source local feature analysis facial
verification program that can be used on a variety of platforms, including embedded
processors, and to determine behavior protocols for the match / non-match cases.

13
 CHAPTER 3
ALGORITHMS

2.1 Algorithms used

Since we use face data as the key here. We use image processing algorithms here. The
implementation is done using OpenCV in python which has built in image recognising
algorithms. There algorithms are trained using machine learning. We use Eigenface
Algorithm

Face recognition is an easy task for humans. Seeing the geometric features of the face is
considered as the most easy and efficient way in facial recognition. On early days the first
automatic face recognition uses the marker points like position of ears, position of nose etc.
These points were used in framing the feature vector like distance between them or angle
between them. The facial recognition process was done by finding the distance between the
feature vector of the captured image with a reference image. Some of the latest work on
geometric face recognition was, a 22-dimensional Threatening people in front of ATM is
done in some places and the security feature of PIN cannot be useful in these situations.

Mainly Computer coding facial recognition can be broken down into three simple stages
that are identical to the steps used by the human brain to recognize faces. These are the
steps:

(1) Data Collection: Gather face data (in this case, face images) of the individuals you want to
recognize.

(2) Train the Recognizer: Feed the recognizer the face data as well as the names of each face
so that it can remember.

(3) Recognition: Feed the face recognizer with new faces of certain individuals to see how it
knows them.

14
 OpenCV has two built-in face recognizers. The names of those face recognizers are:
Eigenfaces and Fisher faces.

2.2 Eigenfaces Face Recognizer Algorithm

In this algorithm, a facial image is a point from a high-dimensional image space and a
lower-dimensional representation is found, where classification becomes easy. The lower-
dimensional subspace is found with Principal Component Analysis (PCA), which identifies
the axes with maximum variance. While this kind of transformation is optimal from a
reconstruction standpoint, it doesn’t take any class labels into account. Imagine a situation
where the variance is generated from external sources, let it be light. The axes with
maximum variance do not necessarily contain any discriminative information at all, hence a
classification becomes impossible.

The functioning of the algorithm is that it considers that not all parts of a face are important
equally to recognize a face. Instead, the main feature like the shape of nose, ears, fore-head
are used and considered how they vary from one another The main idea is to find the region
with maximum difference. Let us consider that there will a drastic amount of variation will
be there when we compare the region of eyes and nose. When multiple faces are to be
compared, the comparison between the faces is done looking at the maximum variation
among the faces, and they help to distinguish the faces. This is the process behind
Eigenfaces recognizer. This recognizer works by looking at all the pictures that are used to
train and extracts the key components that are considered to be relevant and ignores the
other components. Those key components are called as Principal components and these
functions as main source for the recognizer.

This face recognizer also maps the records of the face to the person whose record is that. So,
when a new record needs to be added it follows the following process. The principal
component is extracted first. The comparison process is done next with the images used to
train. Find the image which matches the best. Now map the record with the person name. In
eigenface recognizer the light acts as one of the important needs. Only with the help of light
the, shadowed and illuminated parts of the face is found. These details are used to find an
eigenface which are used to represent a face. The shadows and shapes of two eyes, a nose, a
mouth is the most used parameter in facial recognition.
15
16
 CHAPTER 4

METHODOLOGY

The first and most important step of this project will be to locate a powerful opensource facial
recognition program that uses local feature analysis and that is targeted at facial verification.
This program should be compilable on multiple systems, including Linux and Windows
variants, and should be customizable to the extent of allowing for variations in processing
power of the machines onto which it would be deployed. We will then need to familiarize
ourselves with the internal workings of the program so that we can learn its strengths and
limitations. Simple testing of this program will also need to occur so that we could evaluate
its effectiveness.

Several sample images will be taken of several individuals to be used as test cases – one each
for “account” images, and several each for “live” images, each of which would vary pose,
lighting conditions, and expressions. Once a final program is chosen, we will develop a
simple ATM black box program. This program will server as the theoretical ATM with which
the facial recognition software will interact. It will take in a name and password, and then
look in a folder for an image that is associated with that name. It will then take in an image
from a separate folder of “live” images and use the facial recognition program to generate a
match level between the two.

Finally it will use the match level to decide whether or not to allow “access”, at which point it
will terminate. All of this will be necessary, of course, because we will not have access to an
actual ATM or its software. Both pieces of software will be compiled and run on a Windows
XP and a Linux system. Once they are both functioning properly, they will be tweaked as
much as possible to increase performance (decreasing the time spent matching) and to

17
decrease memory footprint. Following that, the black boxes will be broken into two
components – a server and a client – to be used in a two-machine network.

The client code will act as a user interface, passing all input data to the server code, which
will handle the calls to the facial recognition software, further reducing the memory footprint
and processor load required on the client end. In this sense, the thin client architecture of
many ATMs will be emulated. We will then investigate the process of using the black box
program to control a USB camera attached to the computer to avoid the use of the folder of
“live”images. Lastly, it may be possible to add some sort of DES encryption to the client end
to encrypt the input data and decrypt the output data from the server – knowing that this will
increase the processor load, but better allowing us to gauge the time it takes to process.

18
Most people use an ATM for a lot of purposes and in a lot of scenarios. The most common
three scenarios are considered here and the solutions are found. x The user as himself using
the ATM machine. x If the user is threatened by somebody to get money without the user’s
willing. x If the user’s friend uses the card of the user. First of all, when a user enters the
ATM, two choices are given. Whether the user comes with his card (Number 1) or if the
user’s friend comes with the card (Number 2).

A. User uses the ATM by himself When a user by himself uses the ATM machine, first the
card must be inserted. Now a camera captures the image of the user. Now, the captured
image is compared with the image stored in the database. If both the image is matched
(Eigenface algorithm does the comparison part) then a successfully authenticated
message is displayed. Now the user needs to enter the PIN. If the PIN entered is correct
then further transactions can be done.

B. User is threatened by somebody Now the user is threatened by somebody to get money.
For this scenario we have a robust solution. First the face of the user will match. Now for
entering the right PIN, the user must enter a wrong PIN. This causes a alert at the police
portal which is at the backend. The police portal will be at idle stage if the transactions were
smooth. If the wrong PIN is entered then the alert is received.

19
C. User’s friend user the card When the user’s friend uses the card, obviously the face will
not be matched. For this case, we proposed a clear method. The user should set a PIN and
withdrawal limit in the mobile app. This must be set for each and every transaction. Now the
user’s friend must enter this secret PIN. If the PIN matches, then the amount set by the user
will be automatically withdrawn.

20
fig 1.2 flow chart

21
 Fig 1.3 case 1

Result of case1 The face is authenticated, The PIN is matched. Now the amount is entered.
The process is done successfully.

Fig 1.4 Case 2

22
Result of case 2 if the face is matched and the PIN is entered wrong then an alert is sent at the
police terminal at the back end.

Fig 1.5 Case 3

Result of case 3: If the users friend uses the card, Then the PIN (set by the user) must be
entered correctly After correct entry of the PIN amount is withdrawn.

23
 ADVANTAGES OF PROPOSED SYSTEM

The proposed method uses a biometric, this increases the security feature of the existing
system drastically.

Since we use biometric it is unique for each and every individual.

The condition of threatening the user is well addressed here, and this can result in the safety
of the user and the user’s money.

The withdrawal limit is set if the user’s friend uses the card. This helps the user to be in
control of his account.

Implementation of this method is easy, cheap, and efficient.

24
 CHAPTER

CONCLUSION

On summing up, the method what we proposed is far better in increasing the security feature.
The main goal of our paper is to incorporate the facial recognition feature along with the
existing conventional method for the betterment of the user.

The Eigenface algorithm used here is used for the comparing the face of the user with the
face in the database. Machine learning is used to train (used in OpenCV inbuilt face
recogniser) the face recogniser.

The Adaboost face recognition algorithm has the success rate of 75% but the eigenface
Algorithm produces the success rate of 80%. The main limitation of this system is that it
requires periodic maintenance of the cameras. Twins can be an exception in this system. In
rare cases, photos can be used to bypass the security.

The future scopes of this method are that the use of high-quality durable cameras. 3-d
cameras can be used for the condition of twins and photo bypassing.

25
 REFERNCES

[1] J.J.Patoliya, M.M. Desai, "Face Detection based ATM Security System using Embedded
Linux Platform ", 2nd International Conference for Convergence in Technology (I2CT),
2017.

[2] M.Karovaliyaa, S.Karediab, S.Ozac, Dr.D.R.Kalbande, "Enhanced security for ATM

machine with OTP and Facial recognition features",International Conference on Advanced
Computing Technologies and Applications (ICACTA), 2015.

[3]Sivakumar T. 1 , G. Askok 2 , k. S. Venuprathap, "Design and Implementation of Security

Based ATM theft Monitoring system", International Journal of Engineering Inventions,
Volume 3, Issue 1, 2013.

[4] [4]C. Bhosale, P. Dere, C. Jadhav, "ATM security using face and fingerprint recognition",
Interna- tional Journal of Research in Engineering, Technology and Science, Volume VII,
Special Issue, Feb 2017.

[5] [5]Manoj V , M. Sankar R , Sasipriya S , U. Devi E, Devika T , "Multi Authentication

ATM Theft Prevention Using iBeacon", International Research Journal of Engineering and
Technology (IRJET).

26
 [6] L. Wang,H. Ji, Y. Shi, " Face recognition using maximum local fisher discriminant
analysis",18th IEEE International Conference on Image Processing, 2011

[7] K.Shailaja and Dr.B.Anuradha, "Effective Face Recognition using Deep Learning based
Linear Discriminant Classification ", IEEE International conference on Computational
Interlligence and Computing Research,2016. [8] H. R. Babaei, O. Molalapata and A.H.Y
Akbar Pandor, "Face Recognition Application for Au- tomatic Teller Machines
(ATM)",International Conference on Information and Knowledge
Management(ICIKM),2012.

[9] Chen, Joy Iong Zong. "Smart Security System for Suspicious Activity Detection in
VolatileAreas." Journal of Information Technology 2, no. 01 (2020): 64-72.

[10] Suma, V. "Computer vision for human-machine interactionreview." Journal of trends

inComputer Science and Smart technology (TCSST) 1, no. 02 (2019): 131-139.

[11] Face Recognition using Eigenface algorithm on Laptop camera by Rika Rosnelly,
Mutiara S. Simanjuntak, Ade Clinton sitepu, Mulkan Azhari, Sandy Kosasi in 2020 8th
International Conference on Cyber and IT service Management.

[12] Comparison of the performance of Eigen face and Fisher Face Algorithm in the face
recognition process by Delpiah Wahyuningsih, Chanda Kirana, Rahmat Sulaiman, Hamidah,
Triwanto in 2019 7th International Conference on Cyber and IT service
Management(CITSM).

[13] Face Detection based ATM security system using embedded Linux platform Jignesh J.
Patoliya, Miral M Desai in 2017 2nd International Conference for Convergence in
Technology(I2CT).

27
[14] Real Time Facial Expression and emotion recognition using Eigenfaces, LBPH and
Fisher Algorithms Shrayan Mukhopadhyay, Shilpu Sharma 2020 10th International
Conference on cloud computing, Data Science and Engineering.

[15] Face recognition alert mechanism and idealness for monitoring user motion by Web
Services C. Jayaprakash, V Maheshwari 2016 10th International Conference on Intelligent
Systems and Control(ISCO).

28