Vol 12, Issue 7, July/ 2021

ISSN NO: 0377-9254

AUTOMATIC ATTENDANCE SYSTEM BY FACE
RECOGNITION
1
Dr.CH.Rajendra Babu,2PALLANTI HARINI,3BATTULA KALYANI,4THOKALA
VARSHITA
1,2,,3,4
Computer Science Department, Andhra Loyola Institute Of Engineering And
Technology ,Vijayawada

Abstract: overcome those faults which are having in the existing

systems and provides the characteristics such as
Now-a-days attendance is the major process that
detection of faces, extraction of features, detection of
can be done in any educational institutes and it is a
extracted features, and to analyze and store the student
time taking process for every educator in a piece of
details in a specific file or path. The system integrates
paper. In this project we take the attendance by techniques such as image contrasts, integral images,
recognizing the face of the students and mark the color features and cascading classifier for feature
attendance in a csv file with the date and time so
detection. This project will gives the good accuracy
that we can save the time and the records of student due to the use of features like edge, color, LBP.
data and details of student should be stored in a
particular location or in a particular file. This This project tests for more number of testcases. The
research is aimed to develop less intrusive, cost testcases can be checked in a room to test the accuracy
effective and most efficient as automatic of taking the attendance through face recognition. This
attendance system through face recognition using project comprises for only one student who can sit
Machine learning with Python. Where there are before the camera and also the student must be
the libraries in OpenCv that can perform the registered. But the system takes to store the details of
operations such as detecting, training, testing to get multiple number of students. This can be tested at the
the attendance of a student. In this project face light area conditions and dark area conditions, for
recognition and face detection is used. Where face various facial expressions, for various backgrounds
detection is to detect the face of the student and and also for partial faces.
store it in a dataset after that while taking the
Literature Survey:
attendance it compares the image that stored in a
There are multiple ways to do this project such as
dataset if matches with one of image in a dataset
using a real-time computer vision algorithm in
then the attendance will be recorded.
automatic attendance management system. The
system installed the camera with non-intrusive, which
Keywords: OpenCv(Open source computer vision
can snap images in the classroom and compared the
library, LBP(Linear binary pattern)
extracted face from the image of the camera capturing
with faces inside the system. This system also used
Introduction:
machine learning algorithm which are usually used in
Maintaining and monitoring of attendance play a vital computer vision. But in this project we work on
role in every organization. Automatic attendance HaarCascade classifier which used to extract the
system through face recognition performs like the features of faces. This is mainly a object detection
daily activity of attendance making like a traditional algorithm used to identify faces in an image or a real
process with reducing of person intervention. The time video. The model created from this training is
earlier techniques that failed to store the data in a files available at the OpenCV .These models includes eyes
and also fail in storing of multiple faces or detection detection, face detection etc…
of faces, variations. The proposed system aims to

www.jespublication.com Page No:310

 Vol 12, Issue 7, July/ 2021
ISSN NO: 0377-9254

The main purpose is the features on the image makes it 2)Face detection: It is used to find the face
easy to find out the edges or the lines in the image, or
And face recognition does 2 methods such as:
to pick areas where there is a sudden change in the
intensities of the pixels. The haar feature continuously Verification: It basically compares the input facial
traverses from the top left of the image to the bottom image with the facial image related to the user which is
right to search for the particular feature that means requiring the authentication.
edges traversing. The advantage of the edges feature-
based approaches is to integrate the structural Identification: It basically compares the input facial
information by grouping pixels of face edge map to image with all facial images from a dataset The
line segments. After comparing those pixel working of the project is :
calculations and done the further process.
At first it captures the images of student then store it in
Objectives: a training dataset module when we click on the take
attendance button by using LBP it compares the
The objectives for the project are: integrated image with the original image if it matches
then it marks the attendance otherwise it doesn’t marks
1)Detection of face and store it as sufficient number
or records the attendance.
of images.

2)Extraction of features for the detected face for

further testing process

3)Storing of details of a person in a training dataset

4)Testing of the faces with the trained image

5)Automatic modifications in the files without any

other persons.

Methodology:

The methodology that used in this project is

Fig: System architecture
HaarCascade classifier and LBPH algorithms:

At first it resizes the image where the original image

that we have considered is very large after classifying Test cases:
or resizing of images it loads the image and convert it
into a gray-scale images the reason for this gray Testcases for this project is as follows:
channel is easy to process and it is computationally
less intensive as it contains only one channel of black-
and-white. After converting RGB image to gray we
further classifies or extract features of face using face
classifier and using in built function to detect multiple
faces.

The another algorithm used for this project is

LBP(Linear binary pattern):
Experimental analysis and results:
It actually done 2 methods:

1)Face recognition: It recognizes the facial images

which are already resized and converted to gray scale
image.

www.jespublication.com Page No:311

 Vol 12, Issue 7, July/ 2021
ISSN NO: 0377-9254

After that clicking on take attendance it on the webcam

and captures the face of the person and detects some
time when we click on Q and go to the csv file and
checks there should be recorded the attendance of the
person with the date and time that sits infront of
camera .

This will be the result in the csv file after recording the
attendance

Fig: User interface

When we observe the below figure we can see that it

is a user interface which we can enter the details of the
student for the newly registered and also to take the
attendance for already registered students.

When we click on take images it goes through the

training algorithm and store the images of student in a
training dataset nearly 10 copies with different angles
after storing we have to save the profile otherwise it
cant be save the images that are trained.

Fig: Attendance file

Conclusion:

The attendance with face recognition system is very

effective and very efficient and convenient to maintain
the records in the system rather than in the papers or
books. It reduce the time for teachers and also
maintains the different attendance sheets for different
classes at different time and date.
Fig: Trained images
Acknowledgment:
After saving profile the value at total registrations
updated as well as it displays Profile saved We express our deep sense of Gratitude to
successfully. And the student who have registered our Guide Dr.CH.Rajendra Babu,
those details stored in a csv file with their name and id HOD Department of Computer Science and
which is easy for checking for the lecturers as follows: Engineering (CSE) , Andhra Loyola Institute
of Engineering and
Technology(ALIET),Vijayawada their
unflagging support and continuous
encouragement throughout the project work.
References:
[1]. Automatic Attendance system through face
recognition using machine learning by-
Sumeet Kewalramani
[2]. ” Automatic Student Attendance system
through face recognition using machine
Fig: Student details in csv file learning” by- MR. CHAITANYA P , MR.

www.jespublication.com Page No:312

 Vol 12, Issue 7, July/ 2021
ISSN NO: 0377-9254

SMITHA BHAT, MS. SNEHA R, MS. [5]. AP Gopi (2021), Secure Communication in
SWATI K.S Internet of Things Based on Packet Analysis,
[3]. CR Bharathi, A Naresh, AP Gopi (2021), A Machine Intelligence and Soft Computing,
Node Authentication Model in Wireless 2021.
Sensor Networks With Locked Cluster [6]. A Naresh, PG Arepalli (2021), Traffic
Generation, Design Methodologies and Tools Analysis Using IoT for Improving Secured
for 5G Network, 2021. Communication, Innovations in the Industrial
[4]. D Midhunchakkaravarthy (2021), Internet of Things (IIoT), 2021.
Blockchain embedded congestion control [7]. “Face Recognition: A Literature Survey” by-
model for improving packet delivery rate in W. Zhao, R. Chellappa, P. J. Phillips, A.
ad hoc networks, Machine intelligence and Rosenfeld
soft computing, 2021.

www.jespublication.com Page No:313