You are on page 1of 4

2021 7th International Conference on Advanced Computing & Communication Systems (ICACCS)

Biometrically Secured ATM Vigilance


System
2021 7th International Conference on Advanced Computing and Communication Systems (ICACCS) | 978-1-6654-0521-8/20/$31.00 ©2021 IEEE | DOI: 10.1109/ICACCS51430.2021.9441975

Navin Kumar.M Raghul.S


Department of Electronics and Communication Department of Electronics and Communication
Engineering Engineering
Sri Krishna College Of Technology, Tamilnadu, India Sri Krishna College Of Technology, Tamilnadu, India
navinkumar.m@skct. edu.in 17tuec155@skct. edu. in

Nirmal Prasad. K Naveen Kumar. P


Department of Electronics and Communication Department of Electronics and Communication
Engineering Engineering
Sri Krishna College Of Technology, Tamilnadu, India Sri Krishna College Of Technology, Tamilnadu, India
17tuec13 7@skct. edu. in 17tuec134@skct.edu.in

Abstract— The increase in electronic transactions has facing the issues of crimes and security threats to the existing
contributed to a higher demand for fast and precise card system. The existing ATM system uses PIN and ATM
identification and authentication of users. In ATM s, cards for authentication, which have several drawbacks. To
biometric-based authentication offers different benefits. steal ATM card and their information, criminals use some
Fingerprints and facial identification are used in techniques such as ATM skimming, Cash trapping, Shoulder
biometrical authentication. The drawback of the current surfing, and Card trapping. Some customers use their phone
ATM authentication method is the use of password- numbers, birthdates as their PIN which can be easily guessed
PINs. Since PINs can be tracked and misused quickly. by Fraudsters or hacked by cybercriminals. Biometric
authentication has solutions to these problems of ATM card
Our planned system is intended to provide greater
and PIN. This is because the biological details are unique and
protection for ATMs to achieve security and to resolve
cannot be duplicated by others.
these criminal activities. Here, the PINs are replaced by
randomly created OTP that are sent through the IoT. ATMs have become a high priority target to hackers and
The goal of the work is to fully eradicate the use of ATM stealers. ATMs are susceptible to hacker attacks, fraud,
cards. The customer will be allowed to continue with the robberies, and security breaches. A stable system is needed
transaction after biometric and OTP pin authentication. to detect such anomalous behavior. Now a day’s surveillance
The account will be blocked in the event of three camera has been installed in most of the ATMs. But this can't
consecutive incorrect attempts. This initiative also deals prevent ATM robberies. Some ATM crime occurs due to the
with ATM fraud prevention. In case of any suspicious absence of security guards. Thus, there is a serious need for
an efficient system that can detect suspicious activities and
activity is detected through the vibration sensor results
prevents robbery by catching the crime offender before they
in the closing of ATM doors followed by releasing of the
slip away. Thus, ensures public safety, reduce crime and
fainting gas and alerting the surroundings. This will prevent any serious tragedy.
catch the perpetrator engaged in the crime and prevent
the fraud from taking place. In this paper, the proposed model uses Biometric
authentication which includes fingerprint and face
recognition authentication. The drawback of the prevailing
Keywords— I o T; Biometric authentication; microcontroller; authentication Pattern in ATM is the usage of PIN as
face recognition; fingerprint sensor; OTP (One Time Password); password. Here the PINs are replaced with the randomly
a t m. generated OTP sent through IoT services. In case of any
robbery, the vibrations on the ATM are sensed by the sensor,
I. In t r o d u c t io n
the surrounding w ill be alerted by a buzzer and the door of
An Automated Teller Machine (ATM) was first used in ATM room w ill close immediately and also a fainting gas is
London in 1966. Many theoretical and practical pieces of pumped inside the ATM to make the thief unconscious.
research about ATM have been made throughout the world.
This is because of the increase in the range of ATMs to II. Re l at ed Wo r ks

achieve a cashless economy. The ATM is an electronic The drastic increase in the usage of ATM motivates the
telecommunication device that provides financial researchers to research the development, security, and
transactions such as cash withdrawal, cash deposits, funds enhanced facilities of ATM. Gokul.S, Kukan.S, and
transfer, and payments of utility. Meenakshi. K replaced ATM cards with RFID cards which
ATM fraud has become a global issue that has contain the card number of the customer and instead of
dramatically increased in recent years. ATM fraud has an using the PIN, the fingerprint of the customer was used for
impact on both customers and bank operators. They are authorization [2]. V.Prasanan, R.Sandeep Kumar, and

978-1-6654-0521-8/21/$31.00 ©2021 IEEE 919

Authorized licensed use limited to: California State University Fresno. Downloaded on June 30,2021 at 19:21:01 UTC from IEEE Xplore. Restrictions apply.
2021 7th International Conference on Advanced Computing & Communication Systems (ICACCS)
C.Deepak proposed a system that uses a vibration sensor,
light sensor, smoke sensor, and temperature sensor. I f any
person attempts to damage the ATM or break the ATM, it is
detected employing a vibration sensor and the door is
immediately locked [4]. V.Arunkumar and Vasanth Kumar.
V used the face recognition technique for user authorization.
The system compares the face input of the individuals with
the face details of the users in the database and decides
whether or not a match is found. The system provides a
ranked list of matches by comparing the data of the
individual given with all the other individuals in the
database [1]. Abhinav Muley and Vivek used fingerprint
identification that accelerated the transaction and improved
the level of security. The proposed system replaces the
existing ATM transaction card system with a biometric
fingerprint. Biometrics helps to classify an individual
uniquely with the aid of the person's biological
characteristics [5] and [6] proposed a new approach to
detect Anomalous behaviour in ATM ’s. In [7], Authors gave
a face recognition based new generation ATM machine, FIGURE 1: BLOCK DIAGRAM OF ATM
followed by [8] had iris and fast fingerprint verification and SOFTWARE AND SECURITY SYSTEM HARDWARE
[9] which had added message authentication system.

III. Wo r k in g
previous transaction with that account. After completing a
successful Proceeding, the user can either continue or exit.
We have posited a new concept that enhances the This system has been designed with Python-DBMS along
overall experience, usability, security, and convenience of with the use of hardware Peripherals to provide a cheap
the transaction at the ATM. First, the face and fingerprint of ATM system. Accesses to multiple bank’s accounts are also
the customer should be received and uploaded to the supported in this system.
MySQL database, as shown in the figures 3 and 6. To
achieve interaction with the customer, a user interface is The ATM security system protects the system from
created using Python GUI, as shown in the figure 4. The robbery. It also helps the Cops to catch the Robber without
whole transaction w ill be shown through the user interface. any problems. When anyone tries to damage ATM or break
The image of the face and fingerprint of the customer is into ATM, the vibration sensor detects the vibration. The
coded into digital data and then it gets stored in the vibration sensor triggers an input to the microcontroller.
database. While registering biometric details, the customer Then Microcontroller triggers a buzzer and red LED to alert
w ill be provided with a bank account and balance. They also the surrounding. A servo motor w ill act as an ATM door
want to provide a username and a password while controller, as shown in the figure 8. Along with a buzzer
registering their account to the administrator of the alert, a microcontroller closes the ATM room door by
corresponding bank. triggering the servo motor and also runs a motor pump
which acts as fainting gas releasing pump. To open the
When a customer wants to access ATM, he/she ATM door again; cops want to press the switch that was
firstly place their finger on the fingerprint module. After placed outside of the ATM room. After the door opens the
acceptance, they should show their face to the camera that is green LED w ill be ON to indicate the situation under
before them. Both biometric details are converted into control.
digital code. Now, this digital data is checked against the list
of registered customer’s biometric data. I f data matches, IV. Bio m e t r ic Au t h e n t ic a t io n
then the system w ill send an OTP to the corresponding
account holder’s mobile phone through IoT, as shown in the A. Face recognition
figure 5. I f the provided data not matched, then the system
won't allow the further process. The system asks to enter The first step for many facial-related applications, such as
OTP to the customer. After entering OTP, the system checks face recognition or authentication, is generally face
it with the corresponding data. On breaching three detection. Face detection is a process by which an
reattempts the account w ill be blocked. I f all three individual's identification may be recognized or checked
authentication matches then the customer asks to select a using their face. We use facial embedding, in which each
Bank from the given option. After this selection customer face is turned into a vector and this technique is called deep
allows to make the transaction in the respective Bank metric learning. This method can be separated into three
account. basic stages.

Users can ingress any bank account of his/her, but 1. Face Detection: Detecting faces in the picture is the
it should be linked with their ID. Here customers can do very first task we do. We w ill now get the precise
some of the ATM functions like account details, withdrawal, coordinates of the face for further analysis.
deposit, money transfer. And also, they can view their

920

Authorized licensed use limited to: California State University Fresno. Downloaded on June 30,2021 at 19:21:01 UTC from IEEE Xplore. Restrictions apply.
2021 7th International Conference on Advanced Computing & Communication Systems (ICACCS)

2. Feature Extraction: We cut the face out of the picture.


To pull the features out of the face, we use face
embedding’s here. A neural network takes an image of
the face of the person as input and outputs a vector that
represents a face's most significant features. This vector
is called embedding in machine learning, and so we
refer to this vector as face embedding. The network
learns to generate identical vectors for faces that look
similar when training the neural network. The network
now learns to generate similar vectors for the faces of
the same person.
3. Comparing faces: Now that we have facial embedding’s
for any face in our data stored in the database. The first
step is to use the same network we used above to
measure the face embedding for the image and then
equate this embedding with the remainder of the
embedding we have. I f the embedding produced is
closer or comparable to any other embedding, we FIGURE 3: LOGIN OR REGISTER PAGE
recognize the face.

B. Fingerprint Recognition

The user has to put a finger while registering. The machine


can process the finger pictures and create a finger template
based on the results of the processing, and also store the
template. When matching, the user enters the finger with an
optical sensor and the machine produces a finger template
and compares it with the finger library templates. We can
communicate with the module in a particular format using a
packet of hex codes.

Package Package Package


Header Adder Checksum
Id e n tifie r L e n g th C o n te n t

FIGURE 2: DATA PACKAGE FORMAT FOR


COMMUNICATION FIGURE 4: ACCOUNT MENU PAGE

A Header, Address, Package Identifier, Package Length, These User interfaces are developed using the Python GUI
Package Information, and Checksum are included in the library called tkinter.
data package, which is transmitted to or from the module.
4 :3 3 jR A 93% O
To read the Acknowledge packet, we wrote a function. We
detect fingers and store the image of the detected finger in D e ...

the Image Buffer. Generate a character file from the original


Image Buffer finger image and save it to CharBufferl. Then W e c o u l d n ' t f o u n d G P S d a t a o n t h i s D e v ic e . Y o u c a n
we merge data from the CharBufferl and CharBuffer2 m a n u a lly s e t a lo c a tio n fr o m o u r w e b in te r fa c e

character files and create a prototype that is stored in both


CharBufferl and CharBuffer2. After that, we store the
Demo
specified buffer template (Buffer1/Buffer2) at the assigned
Flash library location. a f e w s e c o n d s a g o - 1 V a r ia b le

V. Re s u l t & D i s c u s s io n
J_____ ^
6 ,5 7 8 a fe w s e c o n d s a g o
Thus, the hardware and the software setups are linked
together and processed for User authentication and security
protocols. Tight security is maintained while the transaction
process, as well as physical security, is also provided for the FIGURE 5: OTP SEND TO CUSTOMER MOBILE
ATM environment, which leads to safe and secure ATM
transactions. OTP are sent to the mobile phones using Ubidots IOT
Development platform, which has android/Mac applications
in their respective app stores.

921

Authorized licensed use limited to: California State University Fresno. Downloaded on June 30,2021 at 19:21:01 UTC from IEEE Xplore. Restrictions apply.
2021 7th International Conference on Advanced Computing & Communication Systems (ICACCS)
VI. Co n c l u s io n

As the result o f fraudulent activities increased in the


card-based system, this proposed ATM system uses
biometric and OTP for authorization. The card-less system
provides more secured cash transactions compared to a card-
based system. The biometric authorizations include both
fingerprint and face recognition, in which the fingerprint of
every individual is uniquely identified. The OTP is sent to
the corresponding account holder's mobile phone through
IOT Services. This combined system provides secure ATM
cash transactions. The vibration detection system detects
the suspicious behaviour o f the person by sensors. This
system helps the Cops to catch the Robber without any
problems.
The proposed system is a prototype o f an ATM system
in which money withdrawal facilities can be added to be
implemented at the ATM. In addition to this, the release of
fainting gas is also shown by a prototype model. By this
means o f three successive identity verification and security
FIGURE 6: FINGERPRINT SENSOR (R307) system, we can provide safety and security to both
customers and bank's money.

Re f e r en c es

[1] Arunkumar. V, Vasanth Kumar. V, Naveenly King. K,


Aravindan.T, “ ATM Security using Face Recognition,”
International Journal Of Current Engineering And Scientific
Research(ijcesr). issn: 2393-8374, (online): 2394-0697,
volume-5, issue-4, 2018.
[2] Gokul.S, Kukan.S, Meenakshi. K, Vishnu Priyan S S, Rolant
Gini J, M.E.Harikumar, “Biometric Based Smart ATM Using
RFID,” Proceedings of the Third International Conference on
Smart Systems and Inventive Technology (ICSSIT 2020)
IEEE Xplore Part Number: CFP20P17-ART; ISBN: 978-1­
7281-5821-1.
[3] Priyanka Hemant Kale and Dr. K. K. Jajulwar, “Design of
Embedded Based Dual Identification ATM Card Security
System,” in 2019 9th International Conference on Emerging
Trends in Engineering and Technology - Signal and
Information Processing (ICETET-SIP-19).
[4] V.Prasanan, R.Sandeep Kumar, C.Deepak, R.Deepak
FIGURE 7: DATABASE FOR STORING USERS FACE Kummar, S.Navin Kumar, “ IOT Based ATM Maintenance
IMAGE and Security System” , International Journal of Applied
After registration and verification, every face image w ill be Engineering Research ISSN 0973-4562 Volume 14, Number
6, 2019 (Special Issue) © Research India Publications.
saved in the database. So, when needed, the bank
[5] Abhinav Muley and Vivek Kute, “Prospective solution to
administrator can access it. bank card system Using fingerprint,” Conference Paper, 2018.
[6] Aditya Parab, Abhishek Nikam, Prajwal Mogaveera, Ashwini
Save, “ A New Approach to Detect Anomalous Behaviour in
ATMs,” 6th International Conference on Advanced
Computing & Communication Systems (ICACCS), 2020.
[7] D.Arun Kumar, B.Iniyan, M.Ahamed Askar, A.Ajay “Face
Recognition Based New Generation ATM Machine,” 5th
International Conference on Advanced Computing &
Communication Systems (ICACCS), 978-1-5386-9533-3/19
@2019 IEEE.
[8] Jameer Basha A., Palanisamv V., Purusothaman T, “Efficient
multimodal biometric authentication using fast fingernrint
verification and enhanced iris features” , Journal of Computer
Science 7 (5): 698-706, 2011 ISSN 1549-3636 © 2011
Science Publications.
[9] Mithun Dutta*, Kangkhita Kaem Psyche, Tania Khatun, Md.
Ashiqul Islam, Md. Azijul Islam, “ ATM Card Security Using
Bio-metric and Message Authentication Technology,” 2018
IEEE International Conference on Computer and
Communication Engineering Technology (CCET).
FIGURE 8: ATM DOOR CLOSED USING SERVO
MOTOR

922

Authorized licensed use limited to: California State University Fresno. Downloaded on June 30,2021 at 19:21:01 UTC from IEEE Xplore. Restrictions apply.

You might also like