2022 1st IEEE International Conference on Industrial Electronics: Developments & Applications (ICIDeA)

Face Recognition based Attendance System

2022 1st IEEE International Conference on Industrial Electronics: Developments & Applications (ICIDeA) | 978-1-6654-2149-2/22/$31.00 ©2022 IEEE | DOI: 10.1109/ICIDEA53933.2022.9970084

Saurabh Singh Rajawat Komal Saxena

Application Development Associate Amity Institute of Information Technology
Accenture, Gurugram, Haryana Amity University, Noida, Uttar Pradesh
[email protected] [email protected]

Abstract – In today’s world, taking student attendance in schools task to maintain attendance. The school and college with the
and colleges is a very significant duty for the faculty. If the faculty total number of students have very difficult work to take
takes attendance physically, it wastes a lot of time. Hence, we attendance manually because there are some chances of
have been taking the help of technology. This will play a vital role student’s proxies and time-consuming by the faculty. The face
in enlightening the excellence of the educational system. Taking
recognition system helps to mark being present of the student
attendance physically takes a lot of time and effort. For the
faculty to maintain and manage the attendance sheet is very by identifying their faces. This system will close the proxies
boring work and it costs time from the lectures. The smart and and human error. Face recognition is one of the best methods
computerized attendance system can help manage attendance. It for proof of the identity of a student. We can implement in the
will help to instigate using several ways of biometric. Face ground of schooling for supervision being present of students.
recognition is unique and the most famous technology nowadays. The face recognition system is separated into multiple phases.
This new technology helps to replace the old traditional sheets. The significant phase in face recognition is to detect the face of
via this model, the problem of forged attendance and proxies can the student. In the first phase for marking the attendance, we
be cracked. In the old system, students just punch the card and have the image of class students, then we can capture the
then don’t go to class. The proposed work that monitors the real-
student faces by the electronic camera which is kept at the
time attendance of the students. The Proposed work defines the
effective algorithm that helps to mark attendance inevitably highest point in the class where the electronic camera will
without human involvement. The most important step in this capture an image without leaving out any student. This camera
system is to detect the faces and recognize them. When we can image is the input of the system. For accurate results, we have
talk about a comparison of the traditional attendance and to perform image enhancement and image acquisition by using
currently proposed system, that we have seen the current system some techniques. To identify the last bench students, we have
help to monitor the student attendance more properly. This to use the histogram equalization of photo desires to be
system will be actual help to maintain the attendance and help completed. In the second phase, we get the image after the
faculty to record the students. image enhancement and image acquisition, and then it will go
for face detection where we can apply some famous
Keywords: - Deep Learning, Automation, Attendance System, algorithms like neural networks, ad-boost algorithms, and
Face Recognition, Image Processing. Online Enrollment system, HAAR algorithms. After applying algorithms, we can detect
Machine Learning Algorithm the faces of the student by these famous algorithms. In the
I. INTRODUCTION third phase, we are working to recognize the face of the student
after detecting his/her face from the image. In face
In school and university, attendance marking is a very recognition, different techniques help to recognize the
important job for faculty. Many schools and universities follow student’s image like PCA, Eigen, and some other algorithm.
the outdated tradition of marking the attendance of the student The best algorithm according to me is Eigen. In this technique
by calling their enrollment number. But now some universities when the student's face is cropped from the image. We have
are applying biometric devices on their campuses such as separated each student's image by using Eigen Features like
punch cards, swipe cards, RFID Cards, and Thumbprint every student has different facial, eyes, nose, and hairs. After
scanners. The school and university now move on towards performing the Eigen algorithm, we compare the generated
Biometric devices such as punch cards for attendance. but the image from the database to recognize the students. So,
enrollment calls out and punch card methods have some comparing with the database we will create the database by
drawback because the enrollment method is done by faculty so enrolling the student. In that database, we can keep the student
its very time consuming and sometimes it makes human information like name, class, enrollment number, his/her
mistakes the other punch card method is not accurate because photo for verification.
anyone can punch anyone card and sometimes there are some
mistakes where present students mark absent or if a student II. LITERATURE SURVEY
loses their punch cards then he/she will get a new card after 3-
4 working days and this is a very difficult time to student to In this author have planned software that built on face
mark attendance. So, I have proposed a new idea for a face Recognition. This system has design in such a way to work
recognition system. The face recognition system is one of the where users can easily understand all functionalities. the
most resourceful methods for marking student attendance. If system has some phase to access by the user [1]. To resolve
there are so many students in the class then it is a very difficult the problem of old manual attendance, the author proposed a
simple and reliable system. No special hardware requirement.

978-1-6654-2149-2/22/$31.00 ©2022 IEEE

95
Authorized licensed use limited to: National Institute of Technology. Downloaded on October 15,2023 at 05:55:56 UTC from IEEE Xplore. Restrictions apply.
He used some algorithms and technology to improve the spaced out then the modification among uninterrupted gray
system performance [2]. In this author have proposed a system levels is more expressive than the gray level to make up your
which helps to take attendance easily. The system takes mind for the power of the human eye. The third thing we can
observation of every student at the entry and exit point. the do is histogram equalization, which improves contrast in
authors focused to upgrade the system in the future by mobile- images. It will work like enlarging out the intensity range of the
based face recognition with proper authorization and a image. Then we have an output image that was generated in
centralized database [3]. The author proposed a smart system, MATLAB after that we consider that output image as a new
that system was implemented into two phases: uncontrolled input image for the second process. Now take the input image
and controlled environment. The methods were used to vary for detecting the faces of every student. The detecting of the
the existing ARS system more effectively, and user Friendly student’s face is done by some face algorithm. Then after
is never wind-up [4]. To resolve the manual attendance detection of the student's face. We will recognize the student
methods, the author proposed a system that performs fittingly faces through each student's face and will be cropped by the
with dissimilar face terminologies, expressions, and posture of output image, after the cropped faces then we will compare
the person. The system has made it movable for simple usage that image from the database of the faces. The database has
smooth when the meetings are on, without disturbing the stored all students' face information and that database is
lecture [5]. This author proposed a system that is a more maintaining all information of the student images. So, we will
reliable and efficient attendance system that system effort for be comparing the output student face with the database student
one-to-many face recognitions at a single period [6]. To faces one by one, after comparing the faces the system will
resolve a manual attendance problem, the author makes a automatically mark the attendance of that student on a server
system that runs on some algorithms which help to distinguish with time, date, and particular subject.
the student’s easy way. He had made a system with no specific
hardware [7]. This author presents a set of experiments on a
difficult face detection dataset. The system results of detection
rate are 94.7 % which is improved by about 18.5 % [8]. This
author proposed a system that helps a divergent combination
of algorithms. This system benefits us to complete desired
results with better exactness and less time consumption [9].
Abhishek proposed a system that helps to save the faculty time
for marking the attendance. He made full software in
MATLAB. He solves the problem of attendance for making
the facial recognition system [10].

III. PROPOSED METHODOLOGY

3.1 ARCHITECTURE
Fig 1. The architecture of the Face Recognition System
This architecture has a better implementation of marking the
student attendance in school and college. This system is also
known as a smart attendance system. It's very helpful for the 3.2 Methodology
teacher to mark the attendance without taking extra time from
the lecture. To implement this system into exertion, we We have some methodologies for implementing the system
essential both computer hardware and software requirements which helps to perform the process. We need some stages are
for our development. To implement this system in the school as follows Stages for Face Recognition Based Attendance
or college, we need a classroom where we will have to require System:
a high-definition camera that we will be secure in the maximum
height where it will capture all students' pictures. When the
camera captures the image of all students, that image becomes
input. The input image is going to be processed by image
processing. First, it will do image acquisition where the image
transforms Real-World Data into an array of numerical data
which could be far ahead wrought on a computer, previously
any image processing can originate an image must be captured
by the camera and transformed into a practicable entity.
Second, we will convert the image into a grayscale image
because a grayscale image has intensity is kept eight-bit Fig 2. Steps to Implement the Face Recognition System
integer giving two hundred fifty-six possible dissimilar shades 3.3 Online Enrollment
of gray from black to white. If the stages are consistently

96
Authorized licensed use limited to: National Institute of Technology. Downloaded on October 15,2023 at 05:55:56 UTC from IEEE Xplore. Restrictions apply.
 3.6 Noise removal and skin clear
The above diagram explains the steps which help to effectively
the marking the student’s attendance. Online Enrollment: Image noise is an unescapable side-effect taking place as an
The first step is enrollment; the student has to enroll in the outcome of image capture. In a camera, if the light which
system. the student has to capture their photo then store that passes through the lens misaligns with the sensors, it will
photo in the database. Students have to store his/her data and generate image noise. even if noise is not so clearly visible in
unique biometric features. This data is store in the different a picture, some kind of image noise is assumed to exist. So, the
packets of each student. Each student to fill a form where effort pictures are taken by the camera, it might hold the noise
he/she will store the student's college details like enrollment, which has to be clean from pictures. the median filter is a solo
name, subject,10-15 photo. The enrollment includes: The first of the most important filters which help to remove noise from
camera will capture the photo of students in the classroom. the image. Skin Clear is a technique, which benefits to growth
Then we will perform an enhanced technique using image the effectiveness of the face detection algorithm. Eigenvector
processing after that enhancing process is completed. We will algorithms are used in our system for the face detection of
do extract the image and store that image into the database that students and it is also increasing the precision if the skin is
database is called the face database. Afterimage store of every classified before the scanning procedure of face images. The
student then each student gets a unique enrollment number. precision of the face algorithm is better-quality after skin
clear.
3.4 Image processing
3.7 Deep Learning
We are using image processing in our project because image
processing helps to get an enhanced image or to citation some Deep learning is nothing but an ordinary example of machine
useful data from the image.it helps to evaluating and learning, more indeed -one of its algorithms.
manipulating the image. We have used MATLAB software
because in MATLAB. We have a pre-defined algorithm and 3.7.1 Face Detection:
package It’s a technology used to discover and classify human faces in
digital images. Face Detection is used algorithms and Machine
3.4.1 Image Acquisition: learning to detect the student's faces in the input images. After
The image acquisition work is to capture the image. First, we the image processing part, we will get an image for face
have to install a high-definition camera in the classroom. The detection. The face detection will distinguish the face of each
camera will be installed at the maximum height in the student from the output pictures. we will be applying some
classroom. So, we can get every student's image on the algorithms to detect the student's faces like Viola and Jones
camera. After capturing the image of each student that image algorithms.
becomes the input image for the system.
3.7.2 Face Recognition:
3.4.2 Image Enhancement: Face Recognition is a technique of categorizing or
authenticating the identity of a specific using their face. Face
It is the process that helps to manipulate the image so that we recognition systems use computer algorithms to preference out
have got a suitable image than the original. Grayscale precise individual details about a student’s face. Face
Conversion: After capturing the image from the camera. recognition is the following stage afterward face detection.
Sometimes that image is not clear or high contrast. So, we have Face recognition can be accomplished by gathering the face as
to manage that image. the grayscale conversion is very simple. of the images. After cropping the image, we have to compare
It’s a color gray in which the red, green, blue colors are easily that image from the face database. that database had been
adjusted or we can say all colors have equal intensity. so, the enrolled by the students. The student’s face is check one by
input image will be converted into a grayscale image. one using Eigen Algorithms.

3.5 Histogram Equalization 3.8 Attendance Marking

Histogram equalization is a technique used to improve image The attendance marking is the last step where students get
quality. It helps to improve the contrast in the image and t h e mark for their class. Completing the verification of all
stretching out the intensity range of the image. the use of student faces effective recognition is completed, the
histogram equalization is removing the contrast of the image attendance had been noticeable on the server.
so we can see the faces of every student in the classroom. It
represents the image in the graphical which helps to distribute
the intensity. It will help to recognize the face of the student.it
generates the histogram of the image which was capture by the
camera after histogram equalization.

97
Authorized licensed use limited to: National Institute of Technology. Downloaded on October 15,2023 at 05:55:56 UTC from IEEE Xplore. Restrictions apply.
 Stage 3: Install the camera at the quiet place where it will

cover all student pictures.

Stage 4: Input the picture taken by the camera.

Stage 5: Image Processing

5.1 Acquired the input picture and perform image

acquisition
5.2 Convert the input picture into a grayscale image after
converting then perform histogram equalization.
5.3 Then remove the noise from the picture, after removing
the noise then perform skin clear, this process
a & b come under image enhancement.

Stage 6: Face Detection

6.1 Crop the student's faces from the input picture which
collect from the upper stage.
6.2 Select the portion of the interest.

Stage 7: Face Recognition

7.1 Equate the cropped pictures with face database pictures.

Spot the attendance-on-attendance server.
7.2 If the picture does not match or there are any new faces,
then go to stage 6 again.

Stage 8: End

V. CONCLUSION
In this proposed system, we have developed a system for
Fig 3. Flow chart of the Face Recognition-based Smart Attendance System using Face Recognition. This paper
Attendance System. presents the effective and precise method of attendance in the
classroom atmosphere that can replace the old traditional
manual methods. The system takings attendance of individual
students by nonstop observation at the entrance and exit
points. Via this model, the probabilities of fake attendance and
IV. ALGORITHM
proxies can be abridged. It protects time and effort, especially
if it is a class with a massive number of students. Similarly, to
The algorithm displays the stage by stage at work of a model. implement this model, not a little specific hardware is required.
For this model, we have been prerequisite to using the A camera device & a single Personal Computer, database
following algorithm. servers are enough aimed at making the smart attendance
system. Current work is attentive to the face detection
ALGORITHM: FACE RECOGNITION FOR
ATTENDANCE SYSTEM. algorithms from images or video frames. We would also try to
implement geolocation so that a person can mark their
attendance using their mobile device, all they need to do is to
Start: Classroom image captured by the camera. scan their face using an application and if they are within the
Result: Attendance marking. location and their faces are recognized then their attendance
would be marked automatically.
PROBLEM EXPLANATION: Identification and Verification
of School and College Student. REFERENCES
Stage 1: Start [1] Abin Abraham, Mehul Bapse, Yash Kalaria, Ahmer Usmani, “Face
Recognition Based Attendance System”, IOSR Journal of Computing
Stage 2: Tell the student to enroll the credential information Engineering, e-ISSN: 2278-0661, p-ISSN: 2278:8727, volume 22, Issue
1, Ser. IV (Jan-Feb 2020), PP 56-60.
and picture store in the face database. [2] Naveed Khan Balcoh, M. Harron Yousaf, Waqar Ahmad, M. Iram Baig,

98
Authorized licensed use limited to: National Institute of Technology. Downloaded on October 15,2023 at 05:55:56 UTC from IEEE Xplore. Restrictions apply.
 “Algorithm for Efficient Attendance Management: Face Recognition [15] P. Anantha Prabha, A. Priya Mahalakshmi, V. Priya, “Random Interim
Base Approach”, IJCSI International Journal of Computer Science Issues, Query and Face Recognition Based Attendance Management System”,
Vol 9, Issue 4, No 1, July 2012 ISSN(Online): 1694-0814. International Journal of Mechanical Engineering, vol. 7, No. 8 August
[3] Nirmalya Kar, Dr. Mrinal Kanti Deb Barma, Ashim Saha,” Study of 2022.
Implementing Automated Attendance System Using Face Recgnition [16] Shreyak Sawhney, Karan Kacker, Samyak Jain, Shailendra Narayan
Technique”, International Journal of Computer and Communication Singh, Rakesh Garg, “Real-Time Smart Attendance System using Face
Engineering, Vol. 1, No. 2, July 2012. Recognition Techniques” 2019 9th International Conference on Cloud
[4] Borra Surekha, Kanchan Jayant Nazare, S. Viswanadha Raju, Nilanjan Computing, Data Science & Engineering (Confluence), 2019, pp 522-
Dey,” Attendance Recording system Using Partial Face Recognition 525.
Algorithm”, Springer International Publishing Switzerland2017. [17] V. Bhalla, T. Singla, A. Gahlot, and V. Gupta, “Bluetooth based
[5] Shubhobrata Bhattacharya, Gowtham Sandeep Nainala, Prosenjit Das, attendance management system,” International Journal of Innovations in
Aurobinda Routray, “Smart Attendance Monitoring System (SAMS): A Engineering and Technology (IJIET) Vol, vol. 3, no. 1, pp. 227–233,
Face Recognition Based Attendance System for Class Environment”, 2013.
2018 IEEE 18th International Conference on Advanced Learning [18] S. S. Mahat and S. Mundhe, “Proposed framework: College attendance
Technologies. management system with mobile phone detector,” International Journal
[6] Priyanka Wagh, Jagruti Chaudhari, Roshani Thakare, Sweta Patil” of Research in IT and Management, vol. 5, no. 11, pp. 72– 82, 2015.
Attendance System Based on face Recognition using Eigen Face and [19] S. Joardar, A. Chatterjee, and A. Rakshit, “A real-time palm dorsa
PCA Algorithms”, 2015 International Conference on Green Computing subcutaneous vein pattern recognition system using collaborative
and Internet of Things (ICGCIoT). representation-based classification,” IEEE Transactions on
[7] Naveed khan Balcoh, M. Haroon Yousaf, Waqar Ahmad, M. Iram Baig, Instrumentation and Measurement, vol. 64, no. 4, pp. 959–966, 2015.
“Algorithm for Efficient Attendance Management: Face Recognition [20] K. Susheel Kumar, Shitala Prasad, Vijay Bhaskar Semwal, R. C.
based Approach”, IJCSI International of Journal of Computer Science Tripathi, "Real Time Face Recognition using AdaBoost Improved Fast
Issues, Vol 9, Issue 4, No 1, July 2012. PCA Algorithm", IJAIA, Vol.2, No. 3, July 2011. A. J. Goldstein, L. D.
[8] Yasaman Heydarzadeh, Abolfazal Toroghi Haghigh at,”An Efficient Harmon, and A. B. Lesk, “Identification of Human Faces,” in Proc. IEEE
Face Detection Method Using Adaboost and Facial Parts”, DOI [21] Sakshi Patel, Prateek Kumar, Shelesh Garg, Ravi Kumar, “Face
10.5013/IJSSST.a.12.04.01 Recognition based smart attendance system using IoT”, JCSE
[9] Divyansh Methi, Abhishek Chauhan, Divyanshu Gupta” Attendance International Journal of computer Sciences and Engineering. Vol-6,
System Using Face Recognition”, International Journal Advanced Issue-5, May 2018.
Research in Science, Engineering and Technology, vol. 4, Issue 5, May [22] Bharath Tej Chinimilli, Anjali T., Akhil Kotturi, Vihas Reddy Kaipu,
2017. Jathin Varma Mandapati,” Face Recognition based Attendance System
[10] Abhishek Jha,”Class Room Attendance System Facial Recognition using Haar Cascade and Local Binary Pattern Histogram Algorithm”,
System”, The International Journal OF Mathematics, Science, 2020 4th International Conference on Trends in Electronics and
Technology and Management vol. 2, issue 3. Informatics (ICOEI)(48184), 2020, pp. 701-704.
[11] Gayatri Gupta, Shweta Jadhav, Vinaya Gurav, Akshay Juwale, Disha [23] P. Vilash, V. Praveen Kumar, K. Srinivas Reddy, N. Sai Kalyan, Mrs
Bhosle, “RFID based attendance System Using Face Recognition”, IOSR Soppari Kavitha, “Automatic Attendance Monitoring System Using
Journal of Engineering (IOSRJEN), Vol.4, PP 23-27. Image Processing”, International Journal of Current Science (IJCPUB)
[12] deAgonia, M. (2017). Apple's Face ID [The iPhone X's facial recognition 2022 IJCPUB, Volume 12, Issue 2 June 2022.
tech explained]. [online] Computerworld. Available at: [24] S. Bhattacharya, G.S. Nainala, P. Das and A. Routray,” Smart Attendance
https://www.computerworld.com/article/3235140/apple-ios/apples-face- Monitoring System (SAMS): A Face Recognition Based Attendance
id-theiphone-xs-facial-recognition-tech-explained.html [Accessed 25 System for Classroom Environment”, 2018 IEEE 18th International
Mar. 2018]. Conference on Advanced Learning Technologies (ICALT), 2018, pp.
[13] Dwi Sunaryono, Joko Siswantoro, Radityo Anggoro, “An android-based 358-360.
course attendance system using Face Recognition”, Journal of King Saud [25] Paras Vishnoi, Priyank Raghav, Manoj Kumar Singh, Pratham
University – Computer and Information Sciences 33(2021) 304-312. Maheshwari, Ranojit Malik,” Smart Attendance System using Face
[14] Chin Howard,” Face Recognition Based Automated Students Attendance Recognition”, International Journal of Advance Engineering Science &
System”, @ 2018, CHIN HOWARD. All right Reserved. Technology, Vol 4, Issue 1, January 2021, pp. 10-14.

99
Authorized licensed use limited to: National Institute of Technology. Downloaded on October 15,2023 at 05:55:56 UTC from IEEE Xplore. Restrictions apply.