Special Issue - 2023 International Journal of Engineering Research & Technology (IJERT)

ISSN: 2278-0181
ICCIDT - 2023 Conference Proceedings

FACIAL RECOGNITION ATTENDANCE

SYSTEM USING FLASK AND OPEN - CV
Mr. Krishnaprasad A R, Ms. Sneha U P , Ms. Sreelakshmi P T
Asst.Prof. Ms. Neena G Krishnan
Dept.of CSE , Mangalam College of Engineering
Ettumanoor, Kerala, India
[email protected], [email protected], [email protected],

Abstract— Based on a person's distinctive facial traits, process. The fundamental tenet is that establishing
face recognition technology employs computer algorithms illumination invariance leads to a feature extraction module
to identify and confirm them. The device employs a that is simpler and has a higher recognition rate. Then, we
camera to take pictures of people, identifies them using used Fuzzy Linear Discriminant Analysis (FLDA) in the
facial recognition software, and records their attendance. feature extraction stage, which demonstrated good
Python and the free computer vision package Open CV discriminating ability in comparison to other methods, and
are used to implement the technique. Face identification Feed forward Neural Network (FFNN) for classification. The
and feature extraction are only two of the many ORL (Olivetti Research Laboratory) face picture database was
techniques for image and video processing that are the subject of the studies, and the findings indicate that the
available in Open CV. Python is a well-liked machine current method is superior to other techniques used on the
learning programming language that is frequently used to same database and described in the literature.
develop and apply facial recognition models. "Face recognition based on fitting a 3D morphable model,"
V. Blanz and T. Vetter, IEEE Trans. on Pattern Analysis and
I. INTRODUCTION Machine Intelligence, vol. 25, no. 9, September 2003 [2]
This project's major goal is to develop a facial recognition- This study proposes a method for face recognition under a
based attendance monitoring system for educational variety of lighting conditions, such as cast shadows and
institutions in order to improve and modernize the current specular reflections, and in poses ranging from frontal to
attendance system and make it more effective and efficient profile views. The method, which estimates 3D shape and
than it was previously. The existing outdated system is very texture of faces from single photos, replicates the process of
ambiguous, which results in erroneous and ineffective image production in 3D space using computer graphics in
attendance taking. When the authority is unable to enforce the order to take these differences into account. A statistical,
rules that are present under the previous system, many issues morphable model of 3D faces is fitted to the photos to provide
occur. The face recognition system will be the technology at the estimate. A collection of textured 3D head scans is used to
work. One of the natural characteristics that can be used to train the model. We go over how the morphable model is
specifically identify a person is their face. The system uses made, an approach for fitting the model to images, and a
facial recognition technology to correctly and swiftly identify framework for face recognition. Faces are represented in this
pupils and mark their attendance, which is an improvement framework by model parameters for 3D shape and texture.
over the time-consuming and error-prone old techniques. "RFID-based Student Attendance System" established by
This project's major goal is to replicate a brand-new, cutting- Hussain, Dugar, Deka, Hannan in 2013 [3]. This paper
edge smart system that can benefit the institution while describes the use of open-source software in a multi-user
resolving the problems with the outdated attendance system. environment to construct a student attendance management
In this project, a program will be created that can recognize system based on radio frequency identification. Python serves
each person's identification and eventually save the as the system's back-end for reading tags. The python
information in a database system. In addition, an excel file is program is run, the lecturers are authenticated, and tags are
made that lists the students' attendance and is sent right away recorded in an XML file using a desktop Java application. The
to the esteemed teachers. XML file is then uploaded to the server to be processed and
used to interpret student attendance. By using the web portal,
users can view attendance.
The phrase "Fingerprint Based Attendance System Using
II. LITERATURE SURVEY
Micro-controller and Lab-view" (2015) Kumar Yadav, Singh,
Pujari, and Mishra [4]. In this context, research prospects are
Face recognition: A convolutional neural-network approach, outlined together with an automated system for recognizing
IEEE Trans. Neural Networks, vol. 8, pp. 98-113, 1997. 2.1 S. fingerprints and the identification of major problems. This
Lawrence, C.L. Giles, A.C. Tsoi, and A.D. Back[1]. In this report's description of an RTOS (Real time operating system)
study, a face recognition system is created that is insensitive implementation in the context of an embedded system reads
to significant variations in lighting. Histogram truncation was like a product design. Despite being very common, fingerprint
utilised as a pre-processing step before the implementation of recognition is a challenging pattern recognition problem. It is
the homomorphic filter in the second step of the normalisation quite challenging to create precise algorithms that can extract

Volume 11, Issue 01 Published by, www.ijert.org

 Special Issue - 2023 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
ICCIDT - 2023 Conference Proceedings

important traits and match them robustly. Here a novel mobile applications and desktop applications that use web
approach is used to simplify the existing problems with a technologies.
proper Embedded System Design . CSS, or cascading style sheets, is an acronym. It is a language
for creating style sheets that govern how HTML texts are
presented and laid out, including the colors, fonts, spacing,
and other visual elements. By separating a document's
appearance from its information using CSS, a web page's
visual style is simpler to manage and change. By setting rules
that are applied to particular HTML elements, CSS operates.
These guidelines outline the preferred design and presentation
of the elements. For instance, you can alter the colour and size
of text, the distance between items, and the layouts for various
screen sizes using CSS.

BACK-END:
Fig.1: Methodology [5]
Python :A high-level, interpreted programming language was
originally made available in 1991. It is a well-liked option
among developers for a variety of applications, including web
III.PROPOSED SYSTEM development, data science, machine learning, and more,
The system will be a web-based application that uses OpenCV because of its simplicity, readability, and applicability.
for face detection and recognition and Flask as the web Python's large ecosystem of third-party libraries and sizable
framework. The two main functions of the system will be standard library make it simpler to complete complicated tasks
adding students to the database and collecting attendance. The with little to no code. Python also supports a variety of
user opens the camera by pressing a button, which then uses programming paradigms, such as procedural, functional, and
haarcascade frontal face detection to take a picture of the object-oriented programming.
student's face and add them to the system. The student's name, Flask Framework: Python-based Flask is a micro web
roll number, and any other pertinent data will then be entered framework that makes it simple and quick for developers to
by the user and stored in a CSV file together with the student's create online apps. It is renowned for being straightforward,
image. The user will click a button to launch the camera and adaptable, and simple to use. Flask is a lightweight, modular
then use OpenCV to detect faces in order to take attendance. framework built on top of the Werkzeug toolkit and the Jinja2
To assess whether the person is a registered student, the template engine.
system will match the detected faces with the pictures in the
database. The student's attendance will be noted in the CSV II. Methodology
file with the date and time if a match is discovered. The front
end of the system will present a table with each student's Before the attendance management system to function, a set of
attendance history and any other pertinent data. The system data that primarily consists of the person's basic information,
will provide straightforward instructions for adding students such as their ID and their faces, must be supplied into the
and taking attendance, making it simple to use. With the right system. The first step in acquiring a portrait is to use the
safeguards in place to make sure that the student's data is not camera to take pictures of people's faces. The system will first
compromised, the system will be secure and preserve the check to see if a face is present in the image that was shot; if it
privacy of the users. isn't, it will encourage the user to take another picture of their
face until the project's minimum need of 10 photographs for
I. Selection of Software each student is met. The decision of storing only 10 portrait
per student is due to the consideration of the limited storage
FRONT-END: space in the raspberry pi because the total amount of students
in the university is considered heavy. Then, the images will
HTML stands for the Hypertext Markup language. It is the undergo several pre-processing procedures to obtain a gray-
standard markup language used to create and design web scale image and cropped faces of equal sized images because
pages. HTML is a set of tags and attributes that define the those are the prerequisites of using the Eigen Faces
structure and content of a web page. HTML tags are used to Recognizer. Both of the processes mentioned above can be
create different types of elements such as headings, represented in Fig.1.
paragraphs, images, links, forms, tables, and more.
HTML documents are text files with the extension ".html" or System hardware requirements
".htm" and are viewed in a web browser. Web browsers Computer: The software components of the attendance
interpret the HTML code and display it as a visual marking system must be operated on a computer. The machine
representation of the web page. HTML is a fundamental should have the minimal requirements needed by the project's
building block of the World Wide Web and is essential for software.
creating static and dynamic web pages, as well as for creating

Volume 11, Issue 01 Published by, www.ijert.org

 Special Issue - 2023 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
ICCIDT - 2023 Conference Proceedings

Camera: A camera is necessary to record student images for

facial recognition. The camera needs to be high-resolution, of
decent quality, and have good lighting skills.
Server: A server may be needed to manage the processing of
facial recognition data as well as to maintain the attendance
records. The server should have enough processing power and
storage to handle the data produced by the system.
Network: In order for the system's various parts, including the
camera, server, and user interface, to communicate with one
another, a network may be necessary. To guarantee that the
data is transmitted and kept safely, the network needs to be
dependable and secure.
Peripherals: A variety of peripherals, such as a printer for
producing attendance records and a bar-code scanner or card Fig.3: Attendance viewing page
reader for identifying students or staff, may be needed for the
attendance marking system. The above figure (Fig.2) refers the admin home page and
Backup tools: To make sure that the attendance records are figure (Fig.3) refers the attendance viewing by admin.
carefully stored and can be easily recovered in case of data Using the User id and password, admin can log on to the
loss, backup tools like external hard drives or cloud storage system and view the attendance.
may be needed. V. CONCLUSION
System software requirements Prior to the creation of this project, the traditional technique of
taking attendance has many flaws, and most institutions have
Operating System - Windows, Linux, IOS experienced significant problems as a result. As a result, the
Programming Language: PYTHON facial recognition technology built into the attendance
Front End : HTML, CSS monitoring system can guarantee accurate attendance capture
Back-End :PYTHON, Flask Web Framework while also removing the drawbacks of the prior approach. By
RAM : Min 4GB handing off the entire complex operation to the machine, using
Processor : i3 and above technology to eliminate flaws not only lowers human
interference in the process but also saves resources. The sole
expense associated with this method is having enough room in
the database storage to accommodate all of the faces.
IV.EXPERIMENT AND RESULTS Thankfully, there are devices like micro SD that can make up
for the data volume. The face database is successfully
constructed in this project. In addition, the face recognition
system is operating effectively. When all is said and done, the
system not only fixes issues with the old model but also makes
it convenient for the user to access the data gathered by
mailing the attendance sheet to the esteemed faculty and
corresponding parents.

VI. REFERENCES
[1] S. Lawrence, C.L. Giles, A.C. Tsoi, and A.D. Back, “Face
recognition: A convolutional neural-network approach,” IEEE
Trans. Neural Networks, vol. 8, pp. 98-113, 1997
[2] V. Blanz and T. Vetter, “Face recognition based on fitting
a 3D morphable model,” IEEE Trans. On Pattern Analysis and
Fig.2: Admin Home page Machine Intelligence, vol. 25, no. 9, September 2003
[3] “RFID based Student Attendance System” (Hussain,
Dugar, Deka, Hannan, 2014)
[4] “Fingerprint Based Attendance System Using Micro-
controller and Lab-view” (Kumar Yadav, Singh, Pujari,
Mishra, 2015)
[5] “Facial Recognition Attendance System Using Python and
OpenCv”( Dr. V Suresh, Srinivasa Chakravarthi Dumpa,
Chiranjeevi Deepak Vankayala,
HaneeshaAduri, Jayasree Rapa,

Volume 11, Issue 01 Published by, www.ijert.org

Special Issue - 2023 International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
ICCIDT - 2023 Conference Proceedings

Volume 11, Issue 01 Published by, www.ijert.org