Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 17

FACIAL DETECTION AND RECOGNITION STUDENT

ATTENDANCE SYSTEM USING IMAGE PROCESSING

TECHNICAL SEMINAR REPORT

Submitted By
AISHWARYA G
(172IT106)

DEPARTMENT OF INFORMATION TECHNOLOGY


BANNARI AMMAN INSTITUTE OF TECHNOLOGY
(AN AUTONOMOUS INSTITUTION AFFILIATED TO
ANNA UNIVERSITY, CHENNAI)
SATHYAMANGALAM-638401
APRIL 2020
BONAFIDE CERTIFICATE

This is to certify that the project entitled “Automated Attendance Using


Facial Recognition” is being submitted by DARUNYA (172IT117). This
is the record of bonafide carried out by the above mentioned student under
our supervision and guidance.

Dr. A. BHARATHI, Mrs. SRINITHYA G


HEAD OF DEPARTMENT, ASSISTANT PROFESSOR,
DEPARTMENT OF INFORMATION TECHNOLOGY DEPARTMENT OF INFORMATION TECHNOLOGY,
BANNARI AMMAN INSTITUTE OF TECHNOLOGY, BANNARI AMMAN INSTITUTE
OF TECHNOLOGY,
SATHYAMANGALAM. SATHYAMANGALAM.

Submitted for Viva Voce examination held on


Internal Examiner External Examiner
ABSTRACT
We are in the world where everything is getting automated. The
internet of things, machine learning, image processing are emerging day
by day. Artificial Intelligence is something which can be defined in simple
terms as a step to feeding the thinking sense of humans to a machine and
making it to perform operations such as detecting, analyzing and
predicting data to automate things. The concept called automated
attendance, where the manual method will be replaced which uses lots of
time. There is so many biometric process, but so far face recognition is the
best one. The teacher need not waste their precious time in taking the
attendance. There is a study which states that the teachers waste up to 3
hours in a month only by taking attendance for a class containing an
average of about 50 students. For a teacher, 3 hours is something which
they wouldn’t like to waste on something which could be automated.
Teachers will come to class only to teach the children and not waste time
on taking attendance. This paper will majorly involve Python, OpenCV
and Dlib to attain the goal.
TABLE OF CONTENTS
CHAPTER NO TITLE PAGE NO

ABSTRACT
INTRODUCTION 1
1.1 Introduction
1.2 Benefits

FACE RECOGNISATION

2 OBJECTIVE 4

3 PROBLEM DEFINITION 5
3.1Adding a Student
3.2Dynamic program
3.3Detecting from the classroom
3.4Marking in an external source
3.5Verification
3.6Backup

4 CONCLUSION 10

5 REFERENCES 11
CHAPTER 1
INTRODUCTION
1.1 INTRODUCTION

The concept of Attendance is now introduced in every place where we


can find employers, students, staffs. After analysing that except the
schools, in all the other places attendance is recorded in a smarter way than
the way it happens at school. In most schools, we have the teachers walk in
and at the beginning of every new hour the teacher is instructed to take an
attendance and mark present for all students that are present in the class.
There was survey which was conducted and it revealed a shocking fact that
almost 3 hours per week is gone just for taking attendance. If any teacher
realizes that they will also be wasting a part of the three hours, they would
end up in disappointment. If there is a way to smarten up the classroom by
automating the attendance which could drastically give the teacher a lot
more time to finish their work. They will have to just walk into the class
and start teaching their subject right away and the attendance for these
hours is scheduled and automated by making use of the CCTV cameras
that will be placed in every classroom. In order to accomplish this mission
there is a need of concept called Artificial Intelligence. These days,
artificial intelligence is a necessity in each and every field that exists. For
an instance let us take road safety, in that if a person is caught violating
any traffic rules then the persons face and the vehicle details are
automatically stored in a database and updated to the police station which
we call as Artificial Intelligent project.

1
Its impact in our society is right there from Siri which is a virtual
assistant to the pacemakers that are placed in a person’s heart to monitor and
work according to that particular person’s heartbeat so as to save him at the
right moment when he needs help. Security and the protection of our society
is drastically improved when artificial intelligence comes into play. Today,
an employer’s salary is calculated using artificial intelligence where during
the date of payment, every minute detail such as his attendance entry using
the biometrics is calculated to see how much the employer’s real salary is.
One major thing when it comes to Facial recognition using Artificial
Intelligence is that there are many ways to implement that and one such way
is using a package called OpenCV in Python. In this particular package, there
are essential library files that will help us to identify and classify objects in a
particular image file. OpenCV is a library filled with packages related to
computer vision which is very useful when it comes to capturing everything
that a camera can capture and analyzed whatever necessary. This analysing
part only changes based on how the users program into python for
manipulating the file that we received as an input either from the camera or
any video that is fed as input to the program.

2
1.2 BENEFITS
The teachers will never have to waste their time on taking
attendance.

They will have more time for completion of their syllabus.

 Student’s will surely get more punctual.

If there is something that goes wrong with the system, there will
also be an alternate way of marking the attendance so that the
students too will not have to worry about marking their
attendance.

3
CHAPTER 2

FACE RECOGNISATION

Face recognition is a task that humans perform routinely and


effortlessly in our daily lives. Wide availability of powerful and low-cost
desktop and embedded computing systems has created an enormous interest
in automatic processing of digital images in a variety of applications,
including biometric authentication, surveillance, human-computer
interaction, and multimedia management. Research and development in
automatic face recognition follows naturally.

OBJECTIVES
To automate the attendance with minimal errors. To make sure that the
teachers get enough time to spend with the students in the classrooms and
for completion of their syllabus.

4
CHAPTER 3
PROBLEM DEFINITION

This paper mainly deals with automating the attendance in a


classroom with the facial details of a student.

This application consists of the following modules.

Adding a Student

Dynamic program

Detecting from the classroom

Marking in an external source

Verification

Backup

5
3.1 ADDING A STUDENT
Adding a student to the database in OpenCV is done by opening the
camera of the device that we will be working on and this will capture around
10-15 frames or photos of a student and store it in a folder. These Pictures
will then be cropped to obtain only the face and after that will be stored under
a unique folder name which will mostly be the unique number of a student.
The pictures inside these will be used for future identification of the student
when they are in their classrooms. Basically, this is a very essential step for
the project since the comparison takes place from the input and the pictures
that are inside these folders and their faces which is stored in the cloud. A
student must compulsorily add their face before this project is implemented
and only then will their attendance be marked and calculated.

6
3.2 DYNAMIC PROGRAM
This is the module which will take care of the program to
execute correctly when an hour begins. The program need not keep on
running but can be restricted using this module where the time that every
hour starts will be noted and given so that this particular detection and
identification process starts exactly at that time and so that these students
receive the attendance for that particular hour.
3.3 DETECTING FROM A CLASSROOM
This module is the heart of the project where detection and
identification of each and every student present in a classroom takes place.
Firstly, this module will start from taking input from the CCTV footage of
the classroom at exactly the mentioned time in the previous module. Once
the start time is encountered, this program will be triggered and it will start
seeking for student’s faces. Once started, it will identify the faces of the
student’s which can be done by the general facial recognition algorithms.
This facial recognition algorithm solely is meant for identifying the faces
present in a given image. Then, after the identification of the face is carried
out, now each face is compared with its respective class folder that
consists of all the student’s facial details of that particular class. This
scanning will go on until all the faces are compared and identified which
will then be kept in a hold state until it is completely done.

7
3.4 MARKING IN AN EXTERNAL SOURCE
Once all the operations such as detecting and identifying, the
values then need to be stored in an external place which is equivalent to
marking the attendance of a student for that hour. In this, we store it in an
external sheet after the identification is done. The final result will be stored
only in this sheet and only from here, a student’s attendance can be
calculated. It will contain some attributes such as every student’s unique
identification number, name of student, ID and their attendance status
where it can be marked as 1(present) or 0(absent). It is also up to the
institution to set up more numbers as options such as 2 for marking that the
student is on leave and 3 to indicate that a student is on duty and has gone
to attend any external event which may be happening off campus.

3.5 VERIFICATION

This is another module can help in improving the results and


accuracy of the program. This is nothing but just to repeat the program for
around 3 times within an hour. This is mainly done so as to improve the
final results. There may be situations where a student’s face might not be
visible at the first time when the attendance is taken. To improve and look
on to such kind of errors, we can improve the result by doing this around
3 times for an hour so that no student misses his attendance

8
3.6 BACK UP
In an artificial intelligence project especially the ones that work
with facial recognition, we can never guarantee the correctness of the
solution that we get from the program. There may be various situations
where a student may not get his attendance due to fail in the facial
recognition module but the student can report to the admins of the
institution and the admins will have the total control over the attendance.
So, if a particular student has a valid reason then his attendance should be
provided. There may be cases where there are power failures where the
CCTV cameras fail, at that particular point of time there should exist a
secondary power source to support them and if there is no proper
secondary power supply then there should also be a provision to manually
take an attendance so that every student gets their attendance.

9
CHAPTER 4

CONCLUSION

In conclusion, this technical seminar mainly deals looks on to the


problem of the amount of time that the teacher wastes on the concept
called attendance which could be automated. If this type of implementation
starts to roll out in our society, we will surely move towards a smarter way
of saving time. The teachers will be very happy to see such a change. All
they will have to do is just walk into the class and start teaching the subject
content that they have prepared for the day without even looking at a
concept called attendance. As always, any solution at the beginning may
not work in a most perfect manner but by implementing it in many
universities, we will surely have feedbacks from all the universities and
then we will get to know every problem that any particular institution will
face and according to that, the solution may be optimized to get a desired
output in such a way that every institution will be satisfied.

10
CHAPTER 5
REFERENCES

[1]Ahmed S. Abdullah, Majida Ali Abed, Israa Al_Barazanchi:


”Improving Face Recognition by elman neural network using curvelet
transform and HSI color space”,Tikrit University,Jan 2019

[2] K. Matsuno, Chil-Woo Lee,S. Kimura,S. Tsuji:”Automatic


recognition of Human facial Recognition”,Cambridge USA,August 2002

[3] Ramesha K, K B Raja, Venugopal K R, L Patnaik:”Feature


Extraction based Face Recognition, Gender and Age Classification”March
2009

11

You might also like