Attendance Management System using Face-Recognition

Authors: Akshit Ramteke, Atharva Shinde, Aditya Sharma, Pir Mohammad

Certificate: View Certificate

Abstract

Nowadays Educational institutions are concerned about regularity of student attendance. Even in a pandemic situation attendance is still a major issue in schools and colleges. Mainly there are two conventional methods of marking attendance which are calling out the roll call or by taking student signs on paper. They both were more time consuming and difficult. Hence, there is a requirement of a computerbased student attendance management system which will assist the faculty for maintaining attendance record automatically. In this project we have implemented the automated attendance system using ‘TKINTER’ and ‘PYTHON’. We have projected our ideas to implement an “Automated Attendance System Based on Face Recognition”. The application includes face identification, which saves time as well as being purely software based; it can be flagged as eco-friendly as it reduces the use of paper. This system also eliminates the chances of fake attendance because of the face being used as a biometric for authentication. Hence, this system can be implemented in a field where attendance plays an important role. The proposed system is designed in TKINTER platform supported with a script of PYTHON as well as SQL database. The algorithm used in the system is based on image comparison based on the encoded values of the face from the image from the database with the image recorded by the system in run time. The system has output in the form of an excel sheet.

Introduction

The Attendance System using Face – Recognition is a
replacement method for the traditional way of marking
attendance. The proposed system is python, a tkinter based
system supported with a MySQL database. This system can
be implemented on a single faculty system of a particular
institute. This system is proposed to be based on biometrics,
i.e., Face Authentication. Since there is the presence of
biometrics, this system eliminates the chances of fake
attendance which is a problem with the traditional methods
of attendance.
The Attendance management is the significant process that
was carried out in every institute to monitor the
performance of the student. Every institute does this in its
own way. Some of their institutes use the old paper or filebased system and some have adopted strategies of
automated attendance systems using some biometric
techniques. A facial recognition system is a computerized
software which is suited for determining or validating a
person by performing comparisons on patterns based on
their facial appearances. In this system OpenCV & Face
Recognition libraries were used which are one of the
popular libraries for face detection by using these libraries
system first capturing the student photos and storing them
into the database which were further used for the training
purpose after that at the time of attendance when system
camera get on system will detect the faces that were present
in the frame the faces were detected by using HOG i.e.
(Histogram of Oriented Gradients) which were carried out
in the system. After that if the image that was present in the
frame is tilted then the Face Landmark Estimation algorithm
will be carried out and the face will be transformed to be as
close as possible to perfectly centered. After that the system
will encode all the images that were present in the database
as well as the faces which were detected in the frame. For
performing encoding Deep Conversional Neural Network
algorithm will be carried out & for each face 128
measurements were generated then the measurements of the
face that were detected in frame it gets compared with the
measurements of the faces that were present in the image
which is earlier stored in the database. So, at last by using a
simple linear SVM algorithm the system will find the
person in the database of known people (i.e., capture at the
starting of the project) who has closest measurements to the
image that were detected by the camera. After finding the
perfect match, the system will generate the name and date &
time & present mark and store the entry in a CSV file.
Which were further uploaded on the database and user can
open it with Microsoft Excel 

Conclusion

In order to obtain the attendance of individuals and to record the entry and exit, the proposed system can be used. The system can widely be used in institutions/organizations. The proposed system takes attendance of each student by continuous observation at the entry and exit points. The result of our preliminary experiment shows improved performance in the estimation of the attendance compared to the traditional attendance marking systems.