IRJET - Face Recognition based Attendance System: Review

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 02 | Feb 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1720
Face Recognition based Attendance System: Review
Neha Savakhande1, Vinay Sripathi2, Kiran Pote3, Parth Shinde4, Prof. Jayant Mahajan5
1,2,3,4Student, Department of Electronics and Telecommunication Engineering, Rajiv Gandhi Institute of Technology,
Mumbai, Maharashtra, India
5Assistant Professor, Department of Electronics and Telecommunication Engineering, Rajiv Gandhi Institute of
Technology, Mumbai, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The planet just where we live in, have taken
many measures to improve the technology. Wherein the
image processing is a cornucopia of innovations which
bolsters the digital and smart systems. When talking about
the models such as identification (ID) of the faces or
generation of attendance, these are few of many practical
applications which are concerned to image processing. This
Paper features and deliberates about the enhancements in
previous and present systems of “Face recognition-based
attendance model” chronologically. The solo purpose of this
research is to critically analyze as well as evaluate the old
and most recent model of “Marking of attendance”. The
conditions and criteria involvedfortheselectionofpapersfor
review are the models which can automatically detect the
pupils in an image captured by camera and mark the
attendance by recognizing the respective students. The
process of attendance is generally known to be taken by
manual methods which is recumbent to security and
maintenance of the attendance log. Plethora of multiple
“Face recognition-based attendance systems” have been
already implemented and there are abundantofissuesstill in
existence. Thus, after studying different attempts made in
researches and a lot of pondering, this paper reviews the
previous works on attendance system based on facial
recognition. It not only providestheliteraturesurveybutalso
supports discussion and suggestions for future work.
Key Words: PCA, LDA, Haar cascade classifier, Face
Recognition, Face detection, Attendance System, viola
jones.
1. INTRODUCTION
The evolution of face identification has received many
vigilances in recent years. It is a centric application in image
analysis, but it is arduous to recreate an automatic model
altogether which has the ability to pinpoint human face [1].
ID’s of the face are vital in the model Management of
Attendance. The non-automatic attendance process is time
snatcher, so a highly qualitative and quantitativeinvestment
of time were given to get liable output. [2] Out of many
answers to this hurdle is the use of model popularly known
as biometric attendance structure. Inevitably it is still
arduous to verify each pupil in a classroom as the volume of
the class is high, and if the model is not able to detect a pupil,
it can disturb the teaching process. There is plethora of
requirements for biometric system or any equivalent model
which are expensive and need a lot of interaction with
students, making it a time snatching model. Due to this, it is
being considered by the researchers has a real challenge for
building a perfect system which can detect and
simultaneously recognize pupils with also a novelty of
marking the presence of the students. The work and
research till now have presented us with faster processing
with adequate amount of efficiency and are also able to
merge it with the technology which deals with Computer
Vision.
Presently, Face identification methods can be found in 2
approaches. Firstly, the method uses localized model which
specifically works on the features such as Nose, Mouth, Eyes
etc. to match an individual face. On the other hand, the other
way is global model. Which considers the whole face when
feature extraction. The above mentioning of these models
has been implemented with the support of many unique
algorithms. The face identificationincludesmanysystematic
proceedings such as, The Image acquisition that deals with
Getting a decent captured image which includes the faces of
the pupils. Followed by extraction which includes detection
of faces to facial feature extractions to the making of
databases. As necessary it sounds, this intermediateprocess
is extremely important because the extraction of feature
should be meaningful so that to further buttress the facial
recognition process. Finally ending the process with Face
matching which consist of face comparison. Albeit it is an
ending operation, but everyprojectcontainspost-processing
which starts after face matching [3].
[4] The operation of face matching can be implemented in
abundant of ways, with the help of many algorithms. The
algorithms which are popularly recognized in helping of
detection of pupils are as follows: - ‘Haar Cascade’, ‘Viola
Jones’ whereas the algorithms which go toe-to-toe with
identification process are as follows: - ‘Eigen Face’, ‘PCA’,
‘Fisher Face’, ‘LDA’, ‘LBPH’. There working may differ and
end results obtained will also be dependent upon the
application i.e. whether it is used for single face recognition
or it is being used for multiple face recognition. The various
algorithms which were and are used in current decade is
been reviewed in the further sections.

2. LITERATURE SURVEY
Tripathi et.al [1] claimed a real timesystemwhich canfollow
through the presence of the students in a classroom. The
necessary supported images for this model was brought ata
constant rate through a webcam until the system is turned
off. The author scanned through several techniques in order
for face detection and encourage them in recognition.Pupils
are distinguished with the help of the Ada boost and Haar
cascade classifier.Althoughforfaceexposerandrecollection,
the author made use of OpenCV libraries butstill forindepth
insight he made a quick use of PCA and LDA. The document
also emphasized about the difference betweenLDAandPCA.
In the end author confidently inclined towards the system’s
accuracy and noted that identification rate is entirely
dependent on the database and the size of the used image.
Ms. Pooja Humbe et.al [2] made use of 360-degree rotating
camera for building the model which detects the pupils in
the class. This system without the software such as XAMPP
controller, NetBeans, Java Advance for the front-
end and back-end with MySQL could have been impossible
as stated by author. The characteristics of face are being
brought by principal component analysis (PCA). Once
registered, the record containing the names of students
attended will be sent through email to parents and teachers.
Shireesha Chintalapati et.al [3] defined the Viola Jones Face
Detection Algorithm. The paper stated that this algorithm
offers better results in various lighting conditions and the
authors have clubbed multiple Haar classifiers to achieve
better output rates up to 30-degree angles. The
preprocessing phase relates to the histogramequalizationof
the facial image obtained in which it is scaled down to
100x100. Images are converted to grayscale; the
equalization of histograms is applied and images are scaled
to size of 100x100. The system employed the LBPH
algorithm to extract the characteristics and the SVM
classifier for classificationpurposeThisdocumentuseda 80-
person database (NITW database) with approximately 20
images of each individual collected for the project. This
document sets out some performance evaluation conditions
when combining LBPH and distance classifier, the false
positive rate is 25 %, the object distance for correct
recognition must be 4 feet, the training time being 563
milliseconds, 95 %of recognition percentage for static
images, the recognition percentage (real-timevideo)was78
%, the occluded faces 2.3% In Microsoft Visual C #and the
EmguCV container the GUI is developedusingtheWinForms
application.
E. Varadharajan et.al [4] explainedtheautomatic Attendance
Management system based on Face Detection. The author
describes how faces are sensed and then cut, before
which background subtraction is performed on the image
in order to improve system performance efficacy. The
erudite authors recommend the use of Eigen face for its
simplicity and quality of performance in facial recognition.
The document also concluded with the observation that in
the case of women, the detection and recognition rate of the
face with a veil was 45% and 10%, while in the case of
women it was 93% and 87% without the veil. The
identification and recognitionlevels,ontheotherhand, were
79% and 65% for bearded men.
Akshara Jadhav et.al [5] prompted face encounter algorithm
Viola Jones and face recognition PCA algorithm withsupport
for machine learning and SVM for extraction functionality.
The author also incorporated reprocessing which includes
the histogram equalization of the facial image extracted and
is scaled to 100x100. The use of neural networks for facial
recognition has been shown, and we can see the possibility
of a semi-supervised learning approach that uses facial
recognition support vector machinesforsatisfactoryresults.
The process followed after the face is recognized is the
subsequent processing in which attendance is generated
weekly or monthly and can be sent to parents or guardians.
Nirmalya Kar et.al [6] used Haar cascade front XML file for
pinpointing a face and confirmationoffacesusing Eigenface.
It was created using Open-CV Libraries. On the end of facial
orientation, the test was prepared. Both detection and
recognition levels were high when facial orientation was
approximately 0 degrees with 98.7% and 95% respectively.
The frequency decreased slowly as facial orientation rose
from 0 degrees to 90 degrees. In the end, the identification
and recognition levels ranged from 0 to 90 degrees.
Smit Hapani et.al [7] has magnified the system which
approbated the model which contributesfacedistinguishing.
Haar classifiers which uses cascade approach and followed
by recognition which uses Fisher face. The system optimally
offers efficacy up to 50% within 15 pupils when modelling
with more than one face with respective to variations such
as cap, spectacles. The proposed system makes use of
classroom through video source, and these resulting frames
are used to identify the faces. Thus, by following the
procedures there by increasing the rate and accuracy of
overall model.
Krishna Dharavath et.al [8] has produced excellent pre-
processing results on a noisy image. The methods suggested
for pre-processing are face cropping,resizing,normalizing &
filtering. A low pass filter is used to eliminate components of
high frequency noise. PCA, DCT (Discrete CosineTransform)
and combined Spatial and Frequency Domain approach are
compared before and after pre-processing. The proposed
combined form has the highest rateoffacerecognitionandis
not much influenced bypre-processing.Themajordrawback
is that facial detection is performed before the pre-
processing of image. In multiplefacerecognitionsystem,this
is not expected as the image needs to be pre-processed first
before any face detection or recognition.

Priyanka Wagh et.al [9] multiple face identification system
has been expressed using Viola Jones for face detection
purposes and the Eigen face forfacerecognition.Itdescribed
the face identification as an invariant of illumination, as it is
a combined form of both Eigen face and PCA. While the face
recognition rate as in the classroom is not established at a
longer distance, the varying lighting conditions do not
impact multiple face recognition.
Nazare Kanchan Jayant et.al [10] executed an automatic
attendance system. This system is based on the Viola Jones
facial detection and face recognition algorithm. First the 20
student’s database is created using various head poses for
culminated recognition results. The face finding algorithm
was then applied, and its efficiency was determined
depending on the number of faces detected. The same
process is followed for calculating the facial recognition
algorithm's efficiency.
Firoz Mahmud et.al [11] approbated use of 2 database types
including UMIST database and ORL database. PCA and LDA
both are used for face knowingpurposes.Theaccuracy ofthe
face recognition is determined using the above listed
algorithms, depending on the face alignment. It is observed
that front aligned faces have a much better accuracy of
recognition than those of face side alignment.
Refik Samet et.al [12] has implemented a fully cell phone
automatic attendance system. This is achieved using the
Viola-Jones algorithm along with Ada-boost trainingforface
finding, since according to the authors, they should work
better in the real-life scenario. For the purposes of
recognition, theEuclidean distance wasdeterminedforthe3
recognition methods, namely its Eigen face, Fisher face and
LBP. A comparison of precision was made for all of the
above-mentioned recognition techniques. The smartphone
application was developed for the automatic attendance
generating system.
Sathyanarayana n et.al [13] launched Automated
Attendance system using facial recognition. The system
specifies algorithms such as Jones' Purple algorithm forface
detection and MSE (medium square error) face recognition.
The document stated and elaborated about the
system’s level of security and accuracy improves as the
number of training images increases. The machine is also
checked for different face angles and alignment up to 60
degrees can be identified. It is observed that when the
system is tested with an image of six students, the system
recognizes five students with 70% efficiency.
D. Nithya et.al [14] has introduced Automated Attendance
System which works on MATLAB. Extractionofthefunctions
is accomplished by analysing the main components. The
Eigen facial approach is utilized for its ease, speed and
learning ability. The difference between the training values
and the test image is calculatedusing the Euclidean distance.
Rajashree P. Suryawanshi et.al [15] prescribed the system
with hardware such asRaspberry PI andawiredcamera,but
the software also consisted of using Open-CV. The very first
step in facial recognition is to detect a face in a given image,
and afterwards proceed withtherecognitiononlyifthereisa
face there. The face pinning was performed using Haar
Cascade Classifier and Face Recognition was based on PCA.
K.L.P.M Liyanage et.al [16] prompted system having a
separate application and a web-based application. The
independent application deals with the process of facial
recognition and the process of marking the attendance. The
Web-based application mainly deals with the NLP process.
Both applications link to a centralized database. Face
detection is achieved using the Haar cascade method while
face recognition is carried out using the PCM method. NLP is
the other research framework developedinSMRT-FRforthe
processing and management of applications for employee
licenses. Employees can easily request authorizations by
sending an SMS or using the web interface and these
requests for authorization are processed using the NLP
application and the result of acceptance or refusal is
generated in the light of different conditions and rules. The
system has been able to detect faces with 68% accuracy so
far.
Professor Arun Katara et.al [17] implemented
a real time assistance system which can perform
multiple facial recognitionusingtheRaspberryPImodel and
Raspberry PI camera. For face pin-pointing it uses Open-CV
libraries and for face recognition the combination of feature
extraction methods such as principal component analysis
along with LBP is implemented. Since the system
can identify faces from a distance of 4 feet to 7 feet,thefacial
recognition efficacy is limited, and is suggested to
be improved. Capture a video of classroom using a video
camera and followed by processing images for
facial recognition.
Kennedy Okokpujie et.al [18] describes a system that uses
Viola Jones as a face detection tool and Fisher face algorithm
for face recognition. Uses a webcam to build the database
and to collect photos to process. It works well in good
lighting conditions, but at different lighting conditions it
decreases the face recognition rate (up to 54%). The system
has access for the authority and the participants via the cell
phone interface with the login credentials.
Nilesh D. Veer et.al [19] an automaticattendancesystem has
been developed in which a video is collected as input.
frames are capturedwhen thereis human presencedetected.
For face detection, Viola Jones is used, and PCA is used for
face recognition, whichalsousesLBPforthreshold purposes.
The facial recognition rate is nearly 100% for a small
number of students and the attendance of the student is
recorded along with the entry time of the student.

A. Majumdar et.al [20] discussed how well they had done
than PCA. To boost dispersion, they used the Fisher face
subspace and LDA, and they also used KNN. Consequently,
better results were obtained by using the Pseudo-
Fisher facial technique. This article examines several
methods that various authorsconsidertoimprovetherateof
detection and recognition. The results show thatViola Jones,
who uses Haar Cascade, is consistent in all the papers
reviewed and offers a good detection rate whereas Fisher
Face's LDA algorithm provides better performance and
faster results.
3. CONCLUSION
This paper reviewed vivid ways in which one can consider
to improve the detection and recognition rate. The results
show that viola jones which uses Haar Cascade is consistent
in all the papers studied and gives a good detection rate
whereas the LDA algorithm which is usedbyfisherfacehasa
better performance and gives faster results. Albeit with the
tedious efforts of authors to adjust those above-mentioned
algorithms with model whichdealswithmultiplefaces, butit
still lacks in both detection and as well as recognition thus
reaching out for Deep Learning with the help of
convolutional neutral networks to engage and satisfy the
need for the application.
REFERENCES
[1] K. Susheel Kumar, Shitala Prasad, Vijay Bhaskar Semwal,
R C Tripathi – “Real time face recognition using adaboost
improved fast pca algorithm” International Journal of
Artificial Intelligence & Applications (IJAIA), Vol.2,No.3,July
2011
[2] Ms. Pooja Humbe, Ms. Shivani Kudale, Ms. Apurva
Kamshetty, Ms. Akshata Jagtap, Prof. Krushna Belerao –
“Automatic Attendance Using Face Recognition”
International Journal on Recent and Innovation Trends in
Computing and Communication Volume: 5 Issue: 12
[3] Shireesha Chintalapati, M.V. Raghunadh – “Automated
Attendance Management System Based OnFaceRecognition
Algorithms” 2013 IEEE International Conference on
Computational Intelligence and Computing Research
[4] E.Varadharajan,R.Dharani, S.Jeevitha, B.Kavinmathi,
S.Hemalatha – “Automatic attendance management system
using face detection” 2016 Online International Conference
on Green Engineering and Technologies (IC-GET)
[5] Akshara Jadhav, Akshay Jadhav Tushar Ladhe, Krishna
Yeolekar – “Automated attendance system using face
recognition” International Research Journal of Engineering
andTechnology (IRJET)
[6] Nirmalya Kar, Mrinal Kanti Debbarma, Ashim Saha, and
Dwijen Rudra Pal “Study of Implementing Automated
Attendance System Using Face Recognition Technique”
International Journal of Computer and Communication
Engineering, Vol. 1, No. 2, July 2012
[7] Smit Hapani, Nikhil Parakhiya, Prof. Nandana Prabhu,
Mayur Paghda – “Automated AttendanceSystemusingImage
Processing” l 2018 Fourth International Conference on
Computing Communication Control and Automation
(ICCUBEA)
[8] Krishna Dharavath,G.Amarnath,Fazal A.Talukdar,Rabul
H. Laskar – “Impact of Image Preprocessing on Face
Recognition: A Comparative Analysis” International
Conference on Communication and Signal Processing (IEEE
2014)
[9] Priyanka Wagh, Roshani Thakare, Shweta Patil, Jagruti
Chaudhari – “Attendance System based on Face Recognition
using Eigen face and PCA Algorithms” 2015 International
Conference on Green Computing and Internet of Things
(ICGCloT) IEEE
[10] Nazare Kanchan Jayant, Surekha Borra – “Attendance
Management System Using Hybrid Face Recognition
Techniques” 2016 Conference on Advances in Signal
Processing (CASP) Cummins College of Engineering for
Women, Pune. Jun 9-11, 2016
[11] Firoz Mahmud, Mst. Taskia Khatun,SyedTauhidZuhori,
Shyla Afroge, Mumu Aktar, Biprodip Pal – “Face Recognition
using PrincipleComponentAnalysisandLinearDiscriminant
Analysis” 2nd Int'l Conf. on Electrical Engineering and
Information & Communication Technology (ICEEICT) 2015
[12] Refik Samet, Muhammed Tanriverdi – “Face
Recognition-Based Mobile AutomaticClassroomAttendance
Management System” 2017 International Conference on
Cyberworlds
[13] Sathyanarayana N,Ramya MR,Ruchitha C,andShwetha
H S – “Automatic Student Attendance Management System
Using Facial Recognition” International Journal of Emerging
Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 25 Issue 6 – MAY 2018.
[14] D. Nithya, Assistant Professor – “Automated Class
Attendance System based on Face Recognition using PCA
Algorithm” International Journal of Engineering Research &
Technology (IJERT) Vol. 4 Issue 12, December-2015
[15] Rajashree P. Suryawanshi1, Vijay Singh Verma2,
Siddarth3, Akhilesh Kumar Singh4 – “Face “recognition
technology based automated attendance system”
International Journal of Advance Engineering and Research
Development
[16] J. G. RoshanTharanga 1*, S. M. S. C. Samarakoon 2, T. A.
P. Karunarathne 2, K. L. P. M. Liyanage 2, M. P. A. W. Gamage2,
D. Perera. 2 (smart - fr) – “Smart attendance using real time
face recognition”

[17] Prof. Arun Katara1, Mr. Sudesh V. Kolhe2, Mr. Amar P.
Zilpe3, Mr. Nikhil D. Bhele4, Mr. Chetan J. Bele – “Attendance
System Using Face Recognition and Class Monitoring
System” International Journal on Recent and Innovation
Trends in Computing and Communication ISSN: 2321-8169
[18] Kennedy Okokpujie, Etinosa Noma-Osaghae, Samuel
John, Kalu-Anyah Grace, Imhade Okokpujie – “A Face
Recognition Attendance SystemwithGSMNotification”2017
IEEE 3rdInternational ConferenceonElectro-Technology for
National Development (NIGERCON)
[19] Nilesh D. Veer, B. F. Momin – “An automated attendance
system using video surveillance camera” IEEE International
Conference On Recent Trends in Electronics Information
Communication Technology, May 20-21, 2016, India
[20] A. Majumdar and R. K. Ward – “Pseudo-Fisherface
method for single image per personfacerecognition”ICASSP
2008
BIOGRAPHIES
PARTH SHINDE
Currently pursuing under
graduation final year in the branch
of EXTC at Rajiv Gandhi instituteof
technology, Versova.
Team leader of the project group.
KIRAN POTE
Team member of the project group
VINAY SRIPATHI
NEHA SAVAKHANDE
nd
Author
Photo
1’st
Author
Photo
Prof. J.R. MAHAJAN(M.E)-project
guide Electronic devices &circuits,
Wireless Communication

IRJET - Face Recognition based Attendance System: Review

More Related Content

What's hot (20)

Similar to IRJET - Face Recognition based Attendance System: Review (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET - Face Recognition based Attendance System: Review