IRJET - Automated Attendance System using Facial Recognition

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2701
Automated Attendance System Using Facial Recognition
Amit S Pawar1, Shubham S Patil2, Kaustubh A Rai3, Roshan Bauskar4
1Amit S Pawar, Dept. of Computer Engineering College, Maharashtra, India
2Shubham S Patil, Dept. of Computer Engineering College, Maharashtra, India
3Kaustubh A Rai, Dept. of Computer Engineering College, Maharashtra, India
4Professor, Roshan Bauskar, Dept. of Computer Engineering, G.V. Acharya Engineering College, Maharashtra, India
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - Facial recognition technologies have undergone
large-scale upgrades and with the new human based facial
recognition algorithm named as ‘Haar-cascade’ [5] which is
used with the three classifiers that are skin classifier, mouth
classifier & eyes classifier [5]. This proposed system works in
real-time as automated attendance system which marks the
attendance of employees in an organization. The proposed
system uses the algorithm of ‘Haar-cascade’ [5] with three
classifiers implemented using python’s ‘Open-CVlibrary’[5] and
this system uses Principle ComponentAnalysis[PCA][1] inorder
to maintain the accuracy of the facial detection..Thisproposed
system requires less hardwaresupportalsotheprocessingtime
is also less in compare to other conventional system of signing
papers, Radio Frequency Identification [RFID] [2] and
biometrics [4] which proves this system to be more efficient for
organization to use it in real time application.
Key Words: Open-CV; Computer Vision; CNN; PCA; LBP;
human face detection; Haar-like features; face skin hue
histogram match; eyes detection; mouth detection;
cascade classifier
1. INTRODUCTION
This project presents the development and implementation
of Smart Attendance System.Oursystemisdesignedinsucha
way that it manages the employee attendance record in a
very efficient manner & time saving pattern that the
employee doesn’t require to fill the attendance sheet or puta
thumb for biometric way or scanning Radio Frequency
Identification [RFID][6] cards instead the Automated
Attendance System concerns about the employee effortsand
doesn’t disturb his work and secretly captures the record of
his/her presence at the time of employee visiting towards
gate and exiting the gate and the record is stored into the
database and which contributes hassle free and costless
attendance of employee. This cost-saving and time-saving
system results a huge profit for organization due to its
efficiency. Our attempt is to develop and implement a Smart
Attendance System for organization in hassle free manner
just by facial detection using Haar-cascade algorithm [1] with
classifier based on skin hue histogram matching, eyes
detection and mouth detection. The organization has to just
place our AutomatedAttendanceSystemintheentrygateand
exit gate of the workingarea of the employees. Wheneverthe
employee will just pass through the gate and automatically
our system will capture his face while during entry period
and exit period and it will store that record intothedatabase.
2. LITERATURE REVIEW
Our proposed system is produced by following all the
previous gadgets and methods for marking and noting down
the attendance. First model we referred was from the paper-
“AttendanceManagementUsingFacialRecognition”fromthis
paper we note that the technologies used in this system was
Principle Component Analysis(PCA), in this systemwefound
that the accuracy is high and less processing time but it has
difficulties to maintain database and it requires continues
power-supply. The second thesis we referred was “Design of
Intelligent Classroom Attendance System Based on Face
Recognition” in this the technologies used were AlexNet
ConvolutionNeuralNetworkcombinedwithRFID,thesystem
was accurate to launch CNN but it wasn’t efficient. Third
thesis we referred was “Smart Attendance Monitoring
System”, in this technology we found that the Viola & Jones
image classification algorithm was usedand the results were
it was accurate and compact but it doesn’t work in areas
where there is insufficient light and also it crashes when
more entries are made into the database. Fourth paper we
referred was “Class Attendance Management System Using
Face Recognition” in this paper the technologies used were
Local Binary Patterns algorithm(LBP) but it also requires
more hardware components. The next system we referred
from thesis “FaceTime- Deep Learning Based Face
Recognition AttendanceSystem”fromthissystemwestudied
that the technologies usedwereConvolutionNeuralNetwork
and CNN cascade, the result it gives is accuracy of 95.02%
with monitoring feature but it takes more time for its
functioning and also fails when there is not stable internet
connectivity. We developedsystemwhichremovesalmostall
the delimitation of referred systems as we use Haar cascade
algorithm with three classifiers and it detects and marks
attendance using face using minimal time & hardware
equipment.
3. PROPOSED SYSTEM
The Proposed system overcomes the problem ofthe existing
system. This project uses the face recognition technique
using the employee records for marking attendance. In the
proposed system application is active during workinghours
of the organization. The camera of the system (Computer
application) will be fitted on entrance as such to scan the
faces of anyone who enters the office, the application
captures the image and sends it to the processing side. The
processing part of the application recognizes the face of the
employee. Finally, the application marks the employee if

he/she is present. If an employee is not recognized by the
application i.e., he / she is not present, he / she is marked as
absent for the day. Humans are prone to error. Automated
attendance management systems ensure accurate time
records and minimize the inevitable and costly errors with
manual data entry. This accurate data thereby helps to
provide accurate performance and payroll data.
Proposed System has the following Significance-
 Monitoring and managing attendance manually can
be a time-consuming, laborious, and expensive
affair. It takes time to process papersheetsandtime
cards, create schedules, authorize leave and
overtime, and create payroll manually.
 The time and effort saved combined with data
accuracy helps in optimizing the use of resources
which lead to increased productivity and improves
profits.
 An integrated attendance management system can
provide good visibility of all data and can ease the
workflow of payrolls, leaves and performance
reviews. Notifications/alerts areautomatedandthe
manager can approve requests for early departure,
overtime, etc., immediately without any specific
need for communication.
 Cloud-based attendance management enables real-
time tracking and provides automated inputs for
payroll processing.
4. SYSTEM ARCHITECTURE
The System process can be separated into three working modules. They are face representation, feature extraction and
classification. The first and foremost task is modeling a face. The way is face is represented determines the next two steps. The
image acquired is transformed to match the positions of images already present.Infeatureextractionthefeaturesofthefaceare
mapped as histograms with gradients and they are stored as binary values. The final step is recognizing a familiar face. The
system compares the face seen in the camera with records that are already stored.
Fig 4.1 -: System Architecture

Fig 4.2 -: Haar like features from OpenCv
Fig -4.3: Flowchart of Proposed System
5. CONCLUSIONS
A new human face detectionalgorithm is proposedonabasis
of cascade classifiers using Haar-like features. Three
additional weak classifiers are subsequentlyappendedtothe
primitive Haar-likefeaturesbasedcascadedclassifiers.Oneis
a decision node based on human skin hue histogram
matching. The secondandthethirdweakclassifiersarebased
on eyes and mouthdetections,respectively.Becauseeyesand
mouth detections are also implemented with Haar-like
features-based cascade classifiers, both of them have a
sufficiently high detection rate, satisfying conditions ofweak
classifiers. Experimental results show that the proposed
human detectionalgorithmcompensatestheshortcomingsof
the primitive Viola-Jones’ cascade classifier and makes the
whole human face detection rate higherwhilekeepingnearly
zero wrong rejection.
The contributions of this work can be concluded as below:
 A weak classifier based on human face skin tone
histogram can reject a big proportion of non-faces
wrongly detected by the primitive Viola-Jones’
Haar-like features-based cascade classifiers.

 2 additional classifiers based on eyes and mouth
detections further remove those non- faces whose
colors happen to be in accordance with the human
skin color, but there are probably no eyes- and
mouth-like objects in it.
 The proposed human face detectionsystemissimple
to implement due to availability of modules in
OpenCV.
REFERENCES
[1] Rajath S Bharadwaj, Tejus S Rao, Vinay T R, “Attendance
Management Using Facial Recognition”, International
Journal of Innovative Technology and Exploring
Engineering (IJITEE) ISSN: 2278-3075, Volume-8Issue-
6, April 2019
[2] Wenxian Zeng, Qinglin Meng, Ran Li, “Design of
Intelligent Classroom Attendance SystemBasedonFace
Recognition”, IEEE 3rd Information Technology,
Networking,
[3] Shubhobrata Bhattacharya, Gowtham Sandeep Nainala,
Prosenjit Das and Aurobinda Routray, “Smart
Attendance Monitoring System (SAMS): A Face
Recognition based Attendance System for Classroom
Environment”, IEEE 18th International Conference on
Advanced Learning Technologies 2018
[4] Omar Abdul Rhman Salim, Rashidah Funke Olanrewaju,
Wasiu Adebayo Balogun“Class AttendanceManagement
System Using Face Recognition”, Department of
Electrical andComputerEngineering, January2018IEEE
[5] Li Cuimei, Qi Zhiliang, Jia Nan, Wu Jianhua, “Human face
detection algorithm via Haar cascade classifier
combined with three additional classifiers”, IEEE 13th
International Conference on Electronic Measurement &
Instruments 2017
[6] Marko Arsenovic, Srdjan Sladojevic, Andras Anderla,
Darko Stefanovic, “FaceTime- DeepLearningBasedFace
RecognitionAttendance System”,IEEE15th International
Symposium on Intelligent Systems & Informatics
September 2017

IRJET - Automated Attendance System using Facial Recognition

More Related Content

What's hot (17)

Similar to IRJET - Automated Attendance System using Facial Recognition (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET - Automated Attendance System using Facial Recognition