Volume 3, Issue 3, March– 2018 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Third Generation ATM using Face Recognition

Mithin P Badiyani1, Aaditya Raj2, T Bharat3, P S Prateek Mohanty4, Preethi V5
1,2,3,4
B.Tech (IV) Year Student , Department of Computer Science and Engineering
SRM Institute of Science and Technology, Chennai, India
5,
Assistant Professor, Department of Computer Science and Technology
SRM Institute of Science and Technology, Chennai, India

Abstract:- Automated teller machines (ATMs) are well Transaction is used because each customer has
acknowledged devices stereotypically used by folks to cart unique fingerprints. So this system provides more secure
out a variability of personal and business economic transaction than GSM. Face recognition method is also used
transactions and banking tasks. ATMs have become very for security in which face is recognized from 3 angles for
popular with the general public for their availability and authentication purpose.
general user friendliness. ATMs are now originate in many
positions having a systematic or high volume of consumer Also, security is enhanced & facial recognition
movement. For example, ATMs are typically found in features. In this system, current face image matched with
restaurants, supermarkets, Convenience stores, malls, stored image and after matching the images correctly, request
schools, gas stations, hotels, work locations, banking will be send to registered number user have to enter the
centers, airports, entertainment establishments, webpage request number for completing transaction. There are
transportation facilities and a myriad of other locations. many systems available for securing transactions, but there is
ATMs are typically available to consumers on a continuous no particular system to secure ATM machine. So there is a
basis such that consumers have the ability to carry out necessity to implement a system which monitors and control
their ATM financial transactions and/or banking functions the room where ATM machines are placed. The design of
at any time of the day and on any day of the week. This proposed system is divided into two parts. The first consists of
based on the facial recognition and also the multilevel a fingerprints reader placed at the entrance of ATM machine
security system based to work this entire concept. Here, we used to identify the user is authorized. The second part
have some PHP support to analyze the person authorized consists of Raspberry Pi module which is placed inside ATM
identification. centre for capturing real-time
video and controlling purpose.
Keywords:- Automated Teller Machine; Face Recognition;
Security; Performance.
II. LITERATURE SURVEY
I. INTRODUCTION A. Design and Implementation ATM Alarm Data Analysis
Nowadays, banking sector is one of the most System
important parts of a human day to day life. Banking facilities Nowadays, people pursuit of fast and convenient way
are widely used by people for their economies activities. of life, fast and convenient service of ATM is made for people
Automatic Teller Machine (ATM) is an electronic machine to avoid waiting in line at the bank for a long time. In order to
which is used for accessing a bank account from anywhere serve people conveniently, it is need to monitor the ATM
without the help of bank staff. The user can perform several equipment to guarantee its normal operation, and deal with the
banking activities like cash withdrawal, money transfer with unexpected problems in time. Therefore, this paper builds a
the help of ATM. It is observed that the number of crimes cloud platform for alarm service, does some alarm analysis,
related to ATM is increased hence there is a necessity to which appears at different times in different locations of the
provide enhances security to ATM machine. Previous ATM machine. This can provide better service for ATM users.
technologies provides security to transactions for identification This system is called ATM Alarm Data Analysis System.
of authorized user. But this is limited for secure transactions
with ATM machine. Previous works focused on biometric B. Data Driven Analytics forAutomated Cell Outage
technique to provide enhanced security to ATM transaction Detection in Self Organizing Networks
whereas GSM based technique is also implemented for the In this paper, we address the challenge of
same purpose. Whereas, some system uses a combination of autonomous cell outage detection (COD) in Self-Organizing
the both techniques. Currently, ATM security is given to the Networks (SON). COD is a pre-requisite to trigger fully
transactions only. GSM based security is provided in which automated self-healing recovery actions following cell outages
One Time Password (WEBPAGE REQUEST) is send to or network failures. A special case of cell outage, referred to
registered number for transaction. The combination of GSM as Sleeping Cell (SC) remains particularly challenging to
and RFID technology is also used which makes the system detect in state-of-the-art SON, since it triggers no alarms for
secure than only RFID technology. This technology has Operation and Maintenance (O&M) entity. Consequently, no
drawback so, biometric technology is introduced for ATM SON compensation function can be launched unless site visits
transaction. In biometric method fingerprint and face or drive tests are performed, or complaints are received by
recognition system are used for ATM transaction. Fingerprint affected customers.
recognition system for ATM

IJISRT18MA441 www.ijisrt.com 655

Volume 3, Issue 3, March– 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
C. New Generation ATM Terminal Services F. Design Theory
This paper aims at realization of the Automated In this system help of the camera will get input image
Teller Machine network all around the globe using IPv6, for face recognition. Before help of PIR will find the person is
thereby reducing the complexity and total number of in front of machine or not after that will enable the camera and
transactions involved in the entire process of cash withdrawal. after first page will open automatically that everything will
But the major challenge involved in connecting ATM network show in monitor. There have select user or third user, if user
to public domain is the security. A Near-Field Communication directly it will move face recognition after recognition done
(NFC) is proposed to be used where in the user, after inserting next our with drawl or if third user is there, third user have to
ATM Card, would communicate via only their NFC enabled update user valid name and password after user will get alert
mobile phones. Reserving NFC spectrum band to the notification via SMS and Mail, based on that user want to give
government is proposed to be made mandatory for ensuring no permission for taking money then third will get money
eavesdropping. otherwise system terminate.

System architecture III. HARDWARE DESCRIPTION

G. Raspberry Pi
The Raspberry Pi is a acclaim card sized single-board
computer with an open-source platform that has a blooming
community of its own, parallel to that of the acoustic. It can be
used in numerous types of projects from novices perception
how to code to design home automation systems. There are a
few versions of the Raspberry Pi, but the latest version, has
improved upon its predecessor in terms of both form and
functionality. The Raspberry Pi Model B features:
 More GPIO
 More USB
 Micro SD
 Lower power consumption
 Better audio
 Neater form factor

Fig 1:- Overview

Here we have the raspberry microcontroller. Inside
that controller we have the some more software use of find
face recognition to the person. We are using the OPENCV
advance image processing software with the help of this we
will find the face of the authorized person. At last we allowing
to take money to that person. What block diagram say know
means here, IR sensor, help of this we will find the persons
entering that room or not, help of PIR sensor find the human in
front of machine or not, at last help of camera take photo and
find the authorized person.
D. Hardware Requirments
 Define IR, PIR sensor.
 CAMERA
 Web camera Fig 2:- Raspberry Pi
 Raspberry pi
The Raspberry Pi is a chains of credit card sized single-board
E. Software Requirements processors developed in the United Kingdom by the Raspberry
Pi Underpinning to indorse the education of
 Raspbian Jessie OS elementary computer skill in schools and mounting countries.
 Language : Python
 Open CV  Raspberry Pi 2 includes a quad-core Cortex-A7 CPU
 HTML, PHP running at 900 MHz and 1 GB RAM. It is described as 4–

IJISRT18MA441 www.ijisrt.com 656

Volume 3, Issue 3, March– 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
6 times more powerful than its predecessor. The GPU is Like shell scripts, Python can computerize errands
identical to the original. like group rechristening and stirring huge volumes of files.
Using IDLE, Python’s REPL (read, eval, print, loop) function
 The Raspberry Pi does not have a built-in real time clock, can be used just like a command line. However, there are more
and does not "know" the time of day. useful things you can create with Python. Programmers use
Python to create things like:
 It have many models All models feature  Web applications
a Broadcom system on a chip (SoC), which includes  Desktop applications and utilities
an ARM compatible central processing unit (CPU) and an  Special GUIs
on chip graphics processing unit (GPU, a Video Core IV).  Small databases
 2D games
 CPU speed runs from 700 MHz to 1.2 GHz for the Pi 3
and on panel recall sort from 256 MB to 1 GB RAM. V. CONCLUSION
Nowadays, most of the ATM has been attacked by
H. PIR Sensor the robberies. In this paper, a real-time monitoring system for
PIR sensors allow you to sense motion, almost ATM security based on accelerometer sensor, camera module,
always used to detect whether a human has moved in or out of and fingerprint module is proposed. The proposed work
the sensors range. They are minor, reasonable, low-clout, concludes with the following points:
relaxed to use and don't garb out. For that reason they are  A secure way of accessing an ATM by authorized persons
normally found in applications and tools used in homes or using face recognition module.
productions. They are often referred to as PIR, "Passive  Eliminates the drawback of previous system like manual
Infrared", "Pyro electric", or "IR motion" sensors. controlling camera modules and doors.
 The system is cost effective as compare to existing
PIRs are fundamentally made of a pyro electronic manual technique.
sensor which can sense ranks of infrared radiation. Everything  The real time video of the ATM centre can be monitored
emits some low level radiation, and the hotter something is, through web server which make ATM better safe from
the more radiation is emitted. The sensor in a motion detector thefts.
is actually split in two halves. The purpose for that is that we
are regarding to spot motion not typical IR levels. The two REFERENCES
halves are wired up so that they cancel each other out. If one [1]. P. Thomas, B. Price, M. Petre, L. Carswell, and M.
half sees more or less IR radiation than the other, the output Richards, “Experiments with electronic examinations over
will swing high or low. the internet,” in Proceedings for the 5th Computer
Assisted Assessment (CAA) Conference, 2001. Available
IV. SOFTWARE DESCRIPTION online: http://caaconference.co.uk/ past
Conferences/2001/proceedings/q1.pdf.
[2]. L. Parmer, “Helping students prepare for qualifying
The Python encoding linguistic essentially initiated as exams; a summary of WCRA Institute III,” in Proceedings
a scripting language for Linux. Python programs are similar to
of the 8th Annual Meeting of the Western College
shell scripts in that the files contain a series of commands that Reading Association, Anaheim, CA, 1975.
the computer executes from top to bottom. Python is a very [3]. S. Görlich, Fundierung und Integration von E-Learning-
beneficial and adaptable high level programming language, Komponenten in die Präsenzlehre (in German), Ph.D.
with tranquil to recite composition that allows computer Thesis, University of Gießen, Faculty 06 – Psychology
operator to use scarcer lines of code than would be probable in and Sports Science, Gießen, Germany, 2006.
languages such as assembly, C, or Java. [4]. W.H. Dotson, J.B. Sheldon, and J.A. Sherman,
“Supporting student learning: improving performance on
Python programs don’t need to be compiled before short-essay exams using realistic practice opportunities,”
running them, as you do with C programs. However, you will Journal of the Scholarship of Teaching and Learning, vol.
need to mount the Python transcriber on computer to route 10(3), pp. 106-118, November 2010.
them. The interpreter is the program that reads the Python [5]. I. Schagaev, N. Folic, N. Ioannides, and E. Bacon,
file and executes the code. There are programs like or r that “Multiple Choice Answers Approach: Assessment with
can package Python code into stand-alone executable penalty function for Computer Science and similar
programs so you can run Python programs on computers disciplines,” International Journal of Engineering
without the Python interpreter installed. Education, vol. 28(6), pp. 1294-1300, 2012.

IJISRT18MA441 www.ijisrt.com 657