Accelerat ing t he world's research.

AUTOMATED ATTENDANCE
SYSTEM USING FACE
RECOGNITION
IRJET Journal

Cite this paper Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers Download a PDF Pack of t he best relat ed papers 

IRJET-AUT OMAT ED AT T ENDANCE SYST EM USING FACE RECOGNIT ION

IRJET Journal

Progression in large Age-gap face verificat ion

IRJET Journal

IRJET-FACE AND FACIAL EXPRESSIONS RECOGNIT ION FOR BLIND PEOPLE

IRJET Journal
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072

AUTOMATED ATTENDANCE SYSTEM USING FACE RECOGNITION

Akshara Jadhav, Akshay Jadhav
Tushar Ladhe, Krishna Yeolekar.
Dept. of Information Technology
University of Pune
NDMVP’s KBT COLLEGE OF ENGINEERING, NASHIK
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In modern times, face recognition has become the whole face to identify a person. The above two
one of the key aspects of computer vision. There are at least approaches have been implemented one way or another by
two reasons for this trend; the first is the commercial and law various algorithms. The recent development of artificial
enforcement applications, and the second is the availability of neural network and its possible applications in face
feasible technologies after years of research. Due to the very recognition systems have attracted many researcher into
nature of the problem, computer scientists, neuroscientists and this field. The intricacy of a face features originate from
psychologists all share a keen interest in this field. In plain continuous changes in the facial features that take place over
words, it is a computer application for automatically time. Regardless of these changes we are able to recognize a
identifying a person from a still image or video frame. In this person very easily. Thus the idea of imitating this skill
paper we proposed an automated attendance management inherent in human beings by machines can be very
system. This system based on face detection and recognition rewarding. Though the idea of developing an intelligent and
algorithms, which automatically detects the student when he self-learning may require supply of sufficient information to
enter in the class room and marks the attendance by the machine. Considering all the above mentioned points and
recognizing him. We used Viola-Jones Algorithm face detection their implications we have tried to gain some experience
which detect human face using cascade classifier and PCA with some of the most commonly available face recognition
algorithm for feature selection and SVM for classification. algorithms and also compare and contrast the use of neural
When compared to traditional attendance marking this network in this field.
system saves the time and also helps to monitor the students.
2. RELATED WORK
Keywords: Face Recognition, Face Detection, Machine
Learning, PCA, SVM. 2.1. RFID (Radio Frequency Identification)-based
Attendance System
1. INTRODUCTION
A number of works related to Radio Frequency Identification
Human beings can distinguish a particular face depending (RFID) based Attendance Systems exist in the literature. In
on a number of factors. One of the main objective of [6] the authors have proposed RFID based system in which
computer vision is to create such a face recognition system students carry a RFID tag type ID card and they need to place
that can emulate and eventually surpass this capability of that on the card reader to record their attendance. RS232 is
humans. In recent years we can see that researches in face used to connect the system to the computer and save the
recognition techniques have gained significant momentum. recorded attendance from the database. This system may
Partly due to the fact that among the available biometric give rise to the problem of fraudulent access. Unauthorized
methods, this is the most unobtrusive. Though it is much person may make use of authorized ID card and enter into
easier to install face recognition system in a large setting, the the organization.
actual implementation is very challenging as it needs to
account for all possible appearance variation caused by 2.2. Fingerprint based Attendance System
change in illumination, facial features, variations in pose,
image resolution, sensor noise, viewing distance, occlusions, In [1] A portable fingerprint device has been developed
etc. Many face recognition algorithms have been developed which can be passed among the students to place their finger
and each has its own strengths. We do face recognition on the sensor during the lecture time without the
almost on a daily basis. Most of the time we look at a face and instructor’s intervention. This system guarantees a fool-
are able to recognize it instantaneously if we are already proof method for marking the attendance. The problem with
familiar with the face. This natural ability if possible imitated this approach is that passing of the device during the lecture
by machines can prove to be invaluable and may provide for time may distract the attention of the students.
very important in real life applications such as various
access control, national and international security and
defense etc. Presently available face detection methods 2.3. Iris-Recognition Based Attendance System
mainly rely on two approaches. The first one is local face
recognition system which uses facial features of a face e.g. Iris is the another bio-metric that can be used for Attendance
nose, mouth, eyes etc. to associate the face with a person. Systems. In [2] the authors have proposed Daugmans
The second approach or global face recognition system use algorithm based Iris recognition system. This system uses

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1467
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072

iris recognition management system that does capturing the 3.3. Pre-Processing
image of iris recognition, extraction, storing and matching.
But the difficulty occurs to lay the transmission lines in the The detected face is extracted and subjected to
places where the topography is bad. In [3] authors have preprocessing. This pre-processing step involves with
proposed a system based on real time face recognition which histogram equalization of the extracted face image and is
is reliable, secure and fast which needs improvement in resized to 100x100. Histogram Equalization is the most
different lighting conditions. common Histogram Normalization technique. This improves
the contrast of the image as it stretches the range of the
3. PROPOSED WORK intensities in an image by making it more clear.

Systems design is the process of defining the architecture, 3.4. Database Development
components, modules, interfaces, and data for a system to
satisfy specified requirements. Systems design could be seen As we chose biometric based system enrolment of every
as the application of systems theory to product development. individual is required. This database development phase
The proposed automated attendance system can be divided consists of image capture of every individual and extracting
into five main modules. The modules and their functions are the bio-metric feature, in our case it is face, and later it is
defined in this section. The five modules into which the enhanced using pre-processing techniques and stored in the
proposed system is divided are: database.

3.1. Image Capture 3.5. Post-Processing

The Camera is mounted at a distance from the entrance to In the proposed system, after recognizing the faces of the
capture the frontal images of the students. And further students, the names are updated into an excel sheet. The
process goes for face detection. excel sheet is generated by exporting mechanism present in
the database system. The database also has the ability to
3.2. Face Detection generate monthly and weekly reports of students attendance
records. These generated records can be sent to parents or
A proper and efficient face detection algorithm always guardians of students. At the end of the class a provision to
enhances the performance of face recognition systems. announce the names of all students who are present in the
Various algorithms are proposed for face detection such as class is also included. This ensures that students whose faces
Face geometry based methods, Feature Invariant methods, are not recognized correctly by the system have the chance
to send a ticket to staff. And thus giving them the ability to
correct the system and make it more stable and accurate. The
announcement system is implemented using text to speech
conversion. Many algorithms and applications are available
that can convert text to lifelike speech. Amazon polly is one
such service which includes 47 lifelike voices spread across
24 languages. Amazon Polly delivers the consistently fast
response times required to support real-time, interactive
dialog. API’s are provided which return the audio stream to
the system. The system also has the ability to send
notification emails to staff and erp operators. The email
would be sent using one of the many Web API’s available like
Amazon SES. The system would use email API’s to send daily
reports to every authorized staff. Push notifications can also
be implemented in the system, notifications can be sent to
Figure 1: System Diagram
both staff and students. Push notifications are real-time and
can help improve the accuracy of the proposed system.
Machine learning based methods. Out of all these methods
Viola and Jones proposed a framework which gives a high
detection rate and is also fast. Viola-Jones detection algorithm 3.6. Proposed Algorithm
is efficient for real time application as it is fast and robust. [9]
Hence we chose Viola-Jones face detection algorithm which 1. Capture the Student’s Image
makes use of Integral Image and AdaBoost learning algorithm 2. Apply Face detection algorithms to detect face
as classier. We observed that this algorithm gives better 3. Extract the Region Of Interest in Rectangular Bounding
results in different lighting conditions and we combined Box
multiple haar classifiers to achieve a better detection rates up 4. Convert to gray scale, apply histogram equalization
to an angle of 30 degrees. and Resize to 100x100 i.e. Apply pre-processing

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1468
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072

5. if Enrollment Phase then to those based on PCA. However, other work such as [5]
Store in Database showed that, when the training data set is small, PCA can
else outperform LDA, and also that PCA is less sensitive to
Apply PCA/LDA/LBPH (For feature Extraction) different training data sets. When applying PCA to a set of
Apply Distance Classifier/SVM/Bayesian (for face images, we are finding a set of basis vectors using lower
Classification) order statistics of the relationships between the pixels.
end if Specifically, we maximize the variance between pixels to
6. Post-processing separate linear dependencies between pixels. MPCA is a
generalization of PCA in that it tries to identify high-order
4. FEATURE SELECTION AND EXTRACTION statistical relationships between pixels to form a better set of
basis vectors where the pixels are treated as random
Throughout the past few decades there have been many face variables and the face images as outcomes. In a similar
detection techniques proposed and implemented. Some of the fashion to PCA and LDA, once the new basis vectors are
common methods described by the researchers of the found, the training and test data are projected into the
respective fields are : subspace and a method such as NN is used for classification.

4.1. Principal component analysis 4.3. Neural Networks

In high-dimensional data, this method is designed to model To model our way of recognizing faces is imitated somewhat
linear variation. Its goal is to find a set of mutually orthogonal by employing neural network. This is accomplished with the
basis functions that capture the directions of maximum aim of developing detection systems that incorporates
variance in the data and for which the coefficients are artificial intelligence for the sake of coming up with a system
pairwise decorrelated [1]. For linearly embedded manifolds, that is intelligent. The use of neural networks for face
PCA is guaranteed to discover the dimensionality of the recognition has been shown by and we can see the suggestion
manifold and produces a compact representation. PCA was of a semi-supervised learning method that uses support
used to describe face images in terms of a set of basis vector machines for face recognition. There have been many
functions, or eigenfaces. Eigenfaces was introduced early [2] efforts in which in addition to the common techniques neural
on as powerful use of principal components analysis (PCA) to networks were implemented. For example in a system was
solve problems in face recognition and detection. PCA is an proposed that uses a combination of eigenfaces and neural
unsupervised technique, so the method does not rely on class network. First The dimensionality of face image is reduced by
information. In our implementation of eigenfaces, we use the the Principal component analysis (PCA) and later the
nearest neighbour (NN) approach to classify our test vectors recognition is done by the Back propagation Neural Network
using the Euclidean distance. Multilinear Principal (BPNN).
Components Analysis One extension of PCA is that of applying
PCA to tensors or multilinear arrays which results in a 5. RESULTS
method known as multilinear principal components analysis
(MPCA) [3]. Since a face image is most naturally a multilinear The attendance management system is simple and works
array, meaning that there are two dimensions describing the efficiently. The project is based on web and needs browser to
location of each pixel in a face image, the idea is to determine access the management system.
a multilinear projection for the image, instead of forming a
one-dimensional (1D) vector from the face image and finding
a linear projection for the vector. It is thought that the
multilinear projection will better capture the correlation
between neighborhood pixels that is otherwise lost in
forming a 1D vector from the image [3].

4.2. Linear Discriminant Analysis

Fisherfaces is the direct use of (Fisher) linear discriminant

analysis (LDA) to face recognition [4]. LDA searches for the
projection axes on which the data points of different classes
are far from each other while requiring data points of the
same class to be close to each other. Unlike PCA which
encodes information in an orthogonal linear space, LDA Figure 2: The home page of management portal.
encodes discriminating information in a linearly separable
space using bases that are not necessarily orthogonal. It is The system has authentication system built in it and needs
generally believed that algorithms based on LDA are superior email and password for access.

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1469
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072

face recognition systems in order to achieve better

performance of the system. Poor lighting conditions may
affect image quality which indirectly degrades system
performance. Our system will perform well but it is not a
perfect solution.

ACKNOWLEDGEMENT

With all respect and gratitude, we would like to thank all the
people who have helped us directly or indirectly for the
completion of the project "Automated Attendance System
Using Face Recognition". We express our heartily gratitude
Figure 3: Sign up page towards Prof. N. S. Ujgare for guiding us to understand the
work conceptually and also for her constant encouragement
The system has two roles student and staff. Both student and to complete the project. Our association with her as a
staff has two different dashboards through which users can student has been extremely inspiring. We would like to give
control and view attendance records. our sincere thanks to Dr. H. V. Patil, Head of the Department
of Information Technology for her technical support and
constant encouragement. We would also like to extend our
sincere thanks to our Principal Dr. K. S. Holkar for his help
and support in all respects. We would also like to thank all
our staff members and collogues who helped us directly or
indirectly throughout our dissertation work.

REFERENCES

[1] B. K. Mohamed and C. Raghu - "Fingerprint attendance

system for classroom needs", in India Conference (INDICON),
2012 Annual IEEE. IEEE, 2012, pp.433438.

Figure 4: Student dashboard [2] T. Lim, S. Sim, and M. Mansor - "Rd based attendance
system", in Industrial Electronics Applications, 2009. ISIEA
2009. IEEE Symposium on, vol. 2. IEEE, 2009,pp. 778782
.
[3] S. Kadry and K. Smaili - "A design and implementation of
a wireless iris recognition attendance management system",
Information Technology and control, vol. 36, no.3, pp.
323329, 2007.

[4] T. A. P. K. K. L. P. M. L. M. P. A. W. G. D. P. J. G..
RoshanTharanga, S. M. S.C. Samarakoon - "Smart attendance
using real time face recognition,", 2013.

Figure 5: Staff dashboard [5] Xiaofei He; Shuicheng Yan; Yuxiao Hu; Niyogi, P.; Hong-
Jiang Zhang, IEEE Transactions on Pattern Analysis and
6. CONCLUSIONS Machine Intelligence, pp. 328 340, 2005.

[6] M. Turk and A. Pentland, Eigenfaces for recognition,

Automated Attendance Systems based on face recognition
Journal of Cognitive Neuroscience, 3(1), pp. 7186, 1991.
techniques thus proved to be time saving and secured. This
system can also be used to identify an unknown person. In
[7] H. Lu, K. N. Plataniotis, and A. N. Venetsanopoulos, Mpca:
real time scenarios PCA outperforms other algorithms with
Multilinear principal component analysis of tensor objects,
better recognition rate and low false positive rate. The future
IEEE Trans. on Neural Networks, 19(1):1839,2008.
work is to improve the recognition rate of algorithms when
there are unintentional changes in a person like tonsuring
[8] Harguess, J., Aggarwal, J.K., A case for the average-half-
head, using scarf, beard. The system developed only
face in 2D and 3D for facerecognition, IEEE Computer Society
recognizes face up to 30 degrees angle variations which has
to be improved further. Gait recognition can be fused with

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1470
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017 www.irjet.net p-ISSN: 2395-0072

Conference on Computer Vision and PatternRecognition

Workshops, pp. 7-12, 2009.

[9] P. N. Belhumeur, J. P. Hespanha, and D. J. Kriegman,

Eigenfaces vs. _sherfaces: Recognition using class speci_c
linear projection, In ECCV 96: Proceedings of the 4th
European Conference on Computer Vision-Volume I, pages
45 58, London, UK,Springer-Verlag 1996.

[10] Martinez, A.M. ; Kak, A.C. , IEEE Transactions on Pattern

Analysis and Machine Intelligence, Volume : 23 , Issue:2, pp.
228 233, Feb 2001.

[11] Shireesha Chintalapati, M.V. Raghunadh , Automated

Attendance Management System Based On Face Recognition
Algorithms, 2013 IEEE International Conference on
Computational Intelligence and Computing Research.

[12] Ming Du, Aswin C. Sankaranarayanan, and Rama

Chellappa , Robust Face Recognition from Multi-View Videos
, IEEE TRANSACTIONS ON IMAGE PROCESSING,VOL. 23, NO.
3, MARCH 2014.

© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1471