See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/337590875

Face Recognition based smart attendance system using IOT

Article in International Journal of Computer Sciences and Engineering · May 2018

DOI: 10.26438/ijcse/v6i5.871877

CITATIONS READS
41 31,053

2 authors, including:

Sakshi Patel
VIT University
14 PUBLICATIONS 173 CITATIONS

SEE PROFILE

All content following this page was uploaded by Sakshi Patel on 28 November 2019.

The user has requested enhancement of the downloaded file.

 International Journal of Computer Sciences and Engineering Open Access
Research Paper Vol.-6, Issue-5, May 2018 E-ISSN: 2347-2693

Face Recognition based smart attendance system using IOT

Sakshi Patel1*, Prateek Kumar2, Shelesh Garg3, Ravi Kumar4
1
Department of ECE, Jaypee University of Engineering and Technology, Madhya Pradesh, India
2
Department of ECE, Jaypee University of Engineering and Technology, Madhya Pradesh, India
3
Department of ECE, Jaypee University of Engineering and Technology, Madhya Pradesh, India
4
Department of ECE, Jaypee University of Engineering and Technology, Madhya Pradesh, India

*Corresponding Author: thesakshipatel@gmail.com, Tel: 8889654455

Available online at: www.ijcseonline.org

Accepted: 16/May/2018, Published: 31/May/2018

Abstract- Attendance is a compulsory requirement of every organization. Maintaining attendance register daily is a difficult and
time consuming task. There are many automated methods for the same available like Biometric, RFID, eye detection, voice
recognition, and many more. This paper provides an efficient and smart method for marking attendance. As it is known that
primary identification for any human is its face, face recognition provides an accurate system which overcomes the ambiguities
like fake attendance, high cost, and time consumption. This system uses face recognizer library for facial recognition and
storing attendance. The absentee’s supervisor or parents are informed through email regarding the absence of their employees
or wards respectively. The objective of this project is to innovate existing projects with some added feature like large data
storage and fast computing through less hardware cost.

Index term –Smart attendance system, Raspberry pi3, OpenCV, Face_Recognition, SMTP

I. INTRODUCTION need of automated attendance system that is fast and reduces

the chance of fake attendance. In this technology system is
Every organization requires an attendance system to
developed for deploying an easy and a secure way of taking
maintain record of presence of student and employees. They
down attendance. This attendance is recorded, by
have their own method to do the same. Some do manually
continuously detecting faces of employees or students via
and some use automated attendance system. Manual method
camera as they enter the classroom. The software first
includes pen and papers which consumes lot of time and
detects the faces and simultaneously compares them with the
wastage of resources. Also, it has risk of proxies and human
predefined database.
error. Automated attendance system includes many methods
like - Section II describes the existing techniques in this field.
Section III of the paper describes the proposed method for
I.I RFID – Radio Frequency Identification is one method for face recognition and generation of attendance report.
attendance making. In this technology an individual has to Section IV gives an idea for hardware used in the proposed
carry his own RFID card. Therefore, this system is cost work. Result is shown in Section V. Conclusion is
effective and can also give rise to fraud as any unauthorized deliberated in Section VI.
person can use the card for fake attendance. II. EXISTING TECHNIQUES

I.II Biometric – This system includes fingerprint, iris and Charles Walton is best known as the first patent holder for
voice recognition. Biometric authentication systems are not the RFID (radio frequency identification) device. He also
100% accurate. False rejection and false acceptance are received a patent for a passive transponder used to unlock a
some errors that can give rise to fraud. door without a key. A card with an
embedded transponder communicated a signal to
I.III Face Recognition – is one of the most efficient systems
a reader near the door. When the reader detected avalid
of all existing ones for identification of people. It can be
identity number stored within the RFID tag, the
used in school, colleges or any organization. To avoid the
reader unlocked the door. [1]
difficulty of taking attendance of large number, there is a

© 2018, IJCSE All Rights Reserved 871

 International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

Michael Dobson, Douglas Ahlers, Bernie DiDario, in 2006

gave a concept of Attendance tracking system. The system Robust 3D Face Recognition: This paper represents a face
includes identification tags, with wireless communication recognition system that overcomes the problem of changes
capabilities, for each potential attendee, scanners for in facial expressions in three-dimensional (3D) range
detecting the attendees' tags as they enter a given room. [2] images. [9]

O. Shoewu and O.A. Idowu: Development of Attendance Monitoring Driver Distraction in Real Time using Computer
Management System using Biometrics. Attendance is taken Vision System: This article presents a real-time non
electronically with the help of a finger print device and the disturbance drowsiness monitoring scheme which exploits
records of attendance are stored in a database. Attendance is the driver’s facial appearance to identify and aware tired
marked after student identification. [3] drivers. This presented work worn the Viola-Jones
Algorithm to identify the driver’s facial appearance. [10]
The idea of using principal components to represent human
faces was developed by Sirovich and Kirby 1987[4] and III. METHODOLOGY
used by Turk and Pentland 1991[5] for face detection and
recognition. It tracks a person’s face and then recognizes by The proposed system uses OpenCV library. It is an Open
comparing characteristics of the face to those of known Source Computer Vision Library that is free for both
individuals. Initially, a principle component factor academic and commercial use. It has C++, Python and Java
“eigenvector” is determined using PCA then the set of interfaces and supports Windows, Linux, MacOS, iOS and
characteristic feature image "eigenface" are found. Android. It has a strong focus on real time applications. The
library has more than 2500 optimized algorithms, these
Paul Viola and Michael Jones in their 2003 article titled,
“Robust Real-Time Face Detection”: It describes a machine algorithms can be used to detect and recognize faces,
learning approach for visual object detection which is identify objects, etc.
capable of processing images extremely rapidly and
achieving high detection rates. It brings together new OpenCV has a Face_Recognizer class library for face
algorithms and insights to construct a framework for robust recognition. This recognizes and manipulates faces from
and extremely rapid object detection. The simple features Python or from the command line. It is a very simple library
used are reminiscent of Haar basis functions which have built using dlib’s state-of-the-art face recognition built with
been used by Papageorgiou et al. deep learning. The dlib is a cross-platform open source
Haar-like features are digital image features used in object
software library that is implemented on multiple computing
recognition. In the detection phase of the Viola–Jones object
detection framework, a window of the target size is moved platforms. The model has an accuracy of 99.38%. This
over the input image, and for each subsection of the image provides a simple face_recognition command line tool that
the Haar-like feature is calculated. This difference is then lets you do face recognition on a folder of images from the
compared to a learned threshold that separates non-objects command line.
from objects. [6]
Fig III.I shows the flow chart of the proposed methodology
LBP (Local Binary Pattern) was first described in 1994 and of the project. The integrated model illustrating the basic
has since been found to be a powerful feature for texture steps for database creation of all the students enrolled in the
classification. It has further been determined that when LBP class, then recognition process for further attendance
is combined with histograms of oriented gradients (HOG) marking, after that comparing the results obtained with the
descriptor, it improves the detection performance predefined database for accurate output, if any student who
considerably on some datasets. It is a centralized approach is present in the database but was absent in the recognition
which involves the step by step process. [7] process will be considered absent in that class. An excel
sheet will be created with attendance record of that day. E-
Automated Attendance Management System Based on Face mail will be send to the absentee parent, informing them the
Recognition Algorithms: This system is based on face attendance for their child.
detection and recognition algorithms, automatically detects
the student when he enters the class room and marks the
attendance by recognizing him. [8]

© 2018, IJCSE All Rights Reserved 872

 International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

Fig: III.I Flow Chart

III.I Database Creation – Database of student is created doing everything you’d expect a desktop computer to do. The
before the recognition process, which includes one image of Raspberry Pi has the ability to interact with the outside world,
each individual. This is done by the respected teacher through and has been used in a wide array of digital maker projects.
the admin panel. During this process admin will
simultaneously enter the student’s name and their parent’s IV.II Webcam - A webcam is a video camera that feeds
email id. After taking images using a camera, an excel sheet or streams its image in real time to or through a computer to
will be created giving the database of all the students. a computer network. When "captured" by the computer, the
video stream may be saved, viewed or sent on to other
III.II Face Recognition – This is the most important part of networks via systems such as the internet, and emailed as an
the proposed system. As students enter the class, teacher will attachment.
switch on the camera and it will continuously detect and
recognize the face. After recognizing all the students present IV.III Display Unit – A two-dimensional display devices
in the class an excel file is created giving the attendance of such as computer monitors and an actual amount
the class with date. of screen space that is available to display a picture, video or
working space.
III.IV Sending Email – As the attendance sheet is ready,
within one minute email will be send to the absentee’s Fig IV.I shows the block diagram of the hardware used in the
employer or parent declaring that their employee or child is model. Main component is Raspberry pi, brain of the system.
absent in today’s job assignment or class. Power is provided using external batteries. Processing, data
creation and result generation are the steps carried out by this
IV. HARDWARE SETUP credit sized computer. A webcam is used to take pictures
during database creation and recognition process. SD card is
Hardware components used in the system are:
connected as external storage, needed for storing data. Using
IV.I RaspberryPi-3 - The Raspberry Pi is a low cost, credit- Ethernet cable, pi is connected to display unit.
card sized computer that plugs into a computer monitor or
Fig IV.II shows the setup and components required for the
TV, and uses a standard keyboard and mouse. It is capable of
proposed model.

© 2018, IJCSE All Rights Reserved 873

 International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

Fig: IV.I Block Diagram

Fig: IV.II Hardware Schematic

V. RESULT Fig V.III: An excel sheet of database record is generated

including every student’s name, image and corresponding
Results obtained by the model are shown in the following
parent’s email id.
figures.
Fig V.IV: Recognition is done for attendance marking.
Fig V.I: Database creation of the students enrolled in the
class. Webcam takes only one image of each student and Fig V.V: An excel sheet of attendance record is generated.
stores for further process.
Fig V.VI: E-mail is send to absentee parent mail using
Fig V.II: During database creation, student’s name and concern faculty id, informing the attendance of their child
parent’s email id is entered. in the particular class.

© 2018, IJCSE All Rights Reserved 874

International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

Fig: V.I Database Creation

Fig: V.II Database created with parent email

Fig: V.III Excel of database created

© 2018, IJCSE All Rights Reserved 875

 International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

Fig: V.IV Recognition

Fig: V.V Attendance record generated

Fig: V.VI E-mail sent to absentee’s parent

VI. CONCLUSION running era. This project consists of modules of OpenCV

algorithms on python platform. This project also allows
As we all know that face is an important identity of each
others to look forward for more improvisation with new
and every person, with the concept of face recognition our
innovation for upcoming changes in hardware &
project describes the importance of automation in the

© 2018, IJCSE All Rights Reserved 876

 International Journal of Computer Sciences and Engineering Vol.6(5), May 2018, E-ISSN: 2347-2693

technology. The model has an accuracy of 99.38% and [10] Monitoring Driver Distraction in Real Time using Computer
provides a simple face_recognition command line tool Vision System. International Journal of Computer Sciences
and Engineering (IJCSE),2017Volume-5, Issue-6, E-ISSN:
which is far more better than the general algorithms as it 2347-2693
needs only one image to work on and not necessary to
convert it in gray scale. Haar cascade, LBPH, eigen face AUTHORS PROFILE

all these need thousands of samples to calculate distance

Sakshi Patel is currently pursing B.Tech
between points and pixels of the image. To make use of degree from Jaypee University of
IOT, email feature is used which is inbuilt in raspberry pi. Engineering and Technology, Madhya
SMT Protocol helps us to do so. The future scope of the Pradesh, India. Her subjects of interests
project can be increasing speed and accuracy of the model. are Wireless Communication and Image
Processing.
ACKNOWLEDGMENT

Authors are highly indebted to JAYPEE UNIVERSITY Prateek Kumar is currently pursing
OF ENGINEERING & TECHNOLOGY GUNA, (MP) for B.Tech degree from Jaypee University
providing the required platform as well as resources of Engineering and Technology,
without which completion of the project would not have Madhya Pradesh, India. His subjects of
interests are Wireless Communication
been. Our deepest thanks to our guide Dr. Ravi Kumar for
and C++ Programming.
taking part in useful decision & giving necessary advices,
guidance and arranged all facilities to make life easier. We
choose this moment to acknowledge his contribution Shelesh Garg is currently pursing
gratefully. B.Tech degree from Jaypee University
of Engineering and Technology,
REFERENCES Madhya Pradesh, India. His subjects of
interests are Embedded and Image
[1] Charles A. Walton, electronic identification & recognition Processing.
system Filed: Dec. 27, 1971 Appl. No.: 212,281.
[2] Michael Dobson, Douglas Ahlers, Bernie DiDario, Ravi Kumar working as Assistant
“Attendance Tracking System", United States Patent Professor(SG) in Department of
Application Publication, Pub. No.: US 2006/0035205 A1,
Electronics and Communication
Feb.16, 2006.
[3] O. Shoewu and O.A. Idowu, “Development of Attendance
Engineering, Jaypee University of
Management System using Biometrics ", The Pacific Engineering & Technology, Guna with
Journal of Science and Technology, Vol. 13, Number1, the specialization in MIMO
pp.300-307, May 2012 (Spring). Communication Systems and Smart
[4] L. Sirovich and M. Kirby (1987). "Low-dimensional Antennas. He has a good research
procedure for the characterization of human faces". Journal exposure and published several research papers in the
of the Optical Society of America A 4: 519–524. various journals of repute.
[5] M.Turk and A. Pentland, "Eigenfaces for Recognition",
Journal of Cognitive Neuroscience, March 1991.
[6] Viola, P., & Jones, M. (2001). Rapid object detection using
a boosted cascade of simple features. In Computer Vision
and Pattern Recognition, 2001. CVPR 2001. Proceedings of
the 2001 IEEE Computer Society Conference on (Vol. 1, pp.
I-511).
[7] C. Papageorgiou, M. Oren, and T. Poggio. A general
framework for object detection. In International Conference
on Computer Vision, 1998.
[8] Automated attendance management system based on face
recognition algorithms Computational Intelligence and
Computing Research (ICCIC), 2013 IEEE International
Conference.
[9] Robust 3D Face Recognition. International Journal of
Computer Sciences and Engineering (IJCSE), 2015 Volume-
3, Issue-5 E-ISSN: 2347-2693

© 2018, IJCSE All Rights Reserved 877

View publication stats