Madan Bhandari Memorial College
New-Baneshowr, Binayaknagar,
Kathmandu,Nepal
Tribhuvan University
Institute of Science and Technology
A Final Year Project Proposal
On
“Facial Recognition Attendance
System”
Submitted to:
Department of Computer Science and
Information Technology
In Partial fulfilment of the requirements for the Bachelor of Science in Computer Science
and Information Technology
Submitted by:
Kushal Bogati(TU Roll No.: 24149/076)
Sanam Rai(TU Roll No.: 24168/076)
Pramod Rawal(TU Roll No.: 24154/076)
Abstract
This project involves building an attendance system which utilizes facial recognition to mark the
presence and time-in of students. It covers areas such as facial detection, alignment, and
recognition, along with the development of a web application to cater to various use cases of the
system such as registration of new students, addition of photos to the training dataset, viewing
attendance reports, etc. This project intends to serve as an efficient substitute for traditional
manual attendance systems. It can be used in corporate offices, schools, and organizations where
security is essential.
ii
Table of Contents
Abstract .......................................................................................................................................... ii
1. Introduction ...................................................................................................................... 1
2. Problem Statement ........................................................................................................... 2
3. Objectives .......................................................................................................................... 3
4. Scope and Limitation ........................................................................................................ 4
5. Research Methodology ..................................................................................................... 5
5.1 Literature Review ........................................................................................................... 6
5.2 Requirements Analysis ................................................................................................... 7
5.3 Feasibility Study ............................................................................................................. 9
5.4 High Level Design and Analysis .................................................................................. 11
6. Technology Used ............................................................................................................. 12
7. Expected Outcome .......................................................................................................... 13
8. References........................................................................................................................ 14
iii
1. Introduction
In today's fast-paced work environment, our project seeks to modernize the way we track
attendance, moving away from the manual methods commonly used. By doing so, we aim to
help organizations maintain digital records of in-time, out-time, break durations, and overall
attendance. This shift towards digitalization not only offers a more efficient alternative to the
traditional attendance system but also brings additional features that enhance its capabilities.
Our primary goal is to simplify attendance tracking by automating the process, reducing the time
and potential errors associated with manual methods. The project also aims to improve record-
keeping accuracy by capturing key details digitally. Moreover, the digital nature of the system
allows for the visualization of data through graphs, providing insights such as the number of
employees present, individual work hours, and break times. This visual representation promises a
more straightforward understanding of attendance patterns and empowers organizations to make
informed decisions.
As an upgrade to existing methods, our project demonstrates a commitment to technological
advancement and operational efficiency. Through the incorporation of advanced features, it not
only tackles current challenges but also sets the stage for a more adaptable and future-ready
attendance management system. In the following sections, we will explore the technical aspects,
functionalities, and anticipated benefits of this innovative solution.
1
2. Problem Statement
Traditional ways of tracking attendance in workplaces are slow, prone to mistakes, and struggle
to adapt to modern work setups. As organizations grow, the manual process becomes inefficient,
leading to errors and delayed access to attendance information. Additionally, these methods face
challenges in accommodating remote and flexible work arrangements, emphasizing the need for
an automated system that aligns with current practices.
Our project tackles these issues by introducing a Facial Recognition Attendance System. The aim
is to replace manual methods with a more accurate, efficient, and flexible solution. We intend to
simplify attendance tracking, overcome the limitations of traditional systems, and improve how
organizations manage attendance in today's diverse work environments.
2
3. Objectives
The project attempts to fulfill the following objectives:
• Automatically record attendance instantly without the need for manual input.
• Simplify the process of registering new students and provide real-time facial image training
with ease.
3
4. Scope and Limitation
Scope:
• Enhance control and efficiency in attendance management, limiting the spread of
misinformation.
• Utilize informational insights for informed predictions about attendance patterns.
• Real-time label extraction for instantaneous and accurate updates to attendance records.
Limitations:
• Challenges with erratic data may lead to anomalies and potential mistakes in attendance
tracking.
• Dependence on weakly labeled data may impact the model's generalization across all cases.
4
5. Research Methodology
Haar cascade
It is a machine learning-based object detection algorithm used to find faces in an image or real-
time video. The algorithm works by comparing rectangular features at different positions in an
image and at different scales. The research article “Rapid Object Detection using a Boosted
Cascade of Simple Features,” written by Paul Viola and Michael Jones in 2001, contains edge or
line detection features that are used by the technique. The algorithm gives a large number of
positive photos with faces and a large number of negative images without any faces to train.
Knn Classifier
KNN algorithm uses the neighborhood classification as the predictive value of a good instance
value. In KNN, the input data is classified based on the closest k data points. For face
recognition, KNN can be used to compare the facial features of an input image to those of the
previously stored image. KNN is also simple to implement and computationally efficient.
5
5.1 Literature Review
[1] Automated Smart Attendance System (2021) - Kolipaka Preethi, Swathy Vodithala
This system proposes a method with stages including Face Detection, DataSet
Creation and Training, and Face Recognition for live attendance marking.
[2] FAREC - CNN Based Face Recognition (2016) - Sharma S, Karthikeyan
Shanmugasundaram, Sathees Kumar Ramasamy
FAREC employs Convolutional Neural Network models, achieving 96% accuracy
for FRGC. The system claims a low False Acceptance Rate (0.1%) and rapid
convergence within 20 epochs.
[3] FaceTime – Deep Learning Based Face Recognition (2017) - Marko Arsenovic,
Srdjan Sladojevic, Andras Anderla, Darko Stefanovic
The model, trained with a small image set and augmentation, shows improved
accuracy. Light conditions during acquisition affect recognition, and the paper
suggests gradient transformation to address issues. Periodic retraining enhances
overall accuracy.
[4] Real-Time Attendance System Using Face Recognition (2020) - Mayank Srivastava,
Amit Kumar, Aditya Dixit, Aman Kumar
The project, using 30 faces for training, focuses on face detection and recognition
rates. The system, intended for colleges, utilizes Paul–Viola's framework. Initial
experiments show better performance than traditional display systems, with a
focus on recognizing faces from images or video frames.
6
5.2 Requirements Analysis
5.2.1 Functional Requirements:
Functional Requirements for Face Recognition Attendance System:
a. User Authentication:
The system should have user authentication mechanisms to ensure secure access to the
application.
b. User Registration:
The system must allow administrators to register users with relevant details, including
facial images for students.
c. Face Enrollment:
The system should allow authorized users to enroll faces for each student during the
registration process.
Captured facial images should be stored securely in the database.
d. Real-Time Processing:
The face detection and recognition process should be performed in real-time to provide
efficient attendance marking during live sessions.
e. Attendance Reports:
Generate attendance reports for specific time periods, classes, or individuals.
Reports should be exportable in common formats (e.g., PDF, CSV).
7
5.2.2 Non-Functional Requirements:
Non- Functional Requirements for Face Recognition Attendance System:
a. Performance:
The system should respond to attendance requests promptly, with minimal latency during
face recognition.
b. Scalability:
The system must be scalable to handle a growing number of users and attendance
records.
c. Security:
Implement robust security measures to protect facial images and attendance data.
Use encryption for data transmission and storage.
d. Accuracy:
The face recognition algorithm should have a high accuracy rate to ensure reliable
attendance marking.
e. Compatibility:
The system should be compatible with various web browsers and devices commonly used
by teachers and administrators.
f. Maintainability:
Design the system in a modular and maintainable way to facilitate updates and
improvements.
8
5.3 Feasibility Study
Feasibility analysis is a crucial step in determining the viability of a proposed project. This
study evaluates the technical, operational, economic, and scheduling aspects of implementing
a Face Detection Attendance System on Django.
5.3.1 Technical Feasibility
The project is technically feasible as it aligns with current technology standards, both in terms
of hardware and software. The following technical requirements are identified:
• A computer with sufficient processing power and memory (e.g., 4GB RAM)
• Webcam or a camera for capturing facial images
• Django framework for web development
• OpenCV library for face detection
• High-speed internet for real-time processing
These technical requirements are readily available and compatible with commonly used
hardware and software components.
5.3.2 Operational Feasibility
The operational feasibility of the project is established by considering the human resources
required. In this case, the project requires a minimum number of personnel. A team of
developers, including expertise in Django and image processing (OpenCV), is deemed
sufficient. The project aims to automate the attendance system through face detection,
minimizing manual effort.
5.3.3 Economic Feasibility
The project is economically feasible due to the utilization of open-source software like
Django and OpenCV, which are freely available. Additionally, many hosting services provide
free hosting up to a certain storage limit, making it cost-effective to deploy and maintain the
system. The benefits derived from automating the attendance system through face detection
outweigh the associated costs, making the project economically viable.
9
5.3.4 Schedule
The estimation of our Project is recorded in Gantt chart below:
10
5.4 High Level Design and Analysis
Fig: Workflow diagram of Facial Recognition Attendance System
11
6. Technology Used
• Language Used:
o Python
• Framework Used:
o Django
• Libraries Used:
o Numpy
o Pandas
o Scikit-learn
o Opencv
12
7. Expected Outcome
In this face recognition-based attendance system, we employ Haar Cascade for face detection
and a K-Nearest Neighbors (KNN) model for recognition. The Haar Cascade classifier is
utilized to identify faces by loading a pre-trained model. Once faces are detected, we
preprocess the images by resizing and converting them to grayscale. Subsequently, a KNN
model is implemented for face recognition, where it is trained with a dataset containing
known faces and their corresponding labels. The attendance system maintains a database of
individuals with unique IDs for tracking attendance. As faces are recognized, the system
marks attendance for the corresponding individuals. Performance metrics, including
accuracy, precision, and recall, are calculated to evaluate the system's effectiveness. The
output displays the names of recognized individuals and indicates whether the face matches
the database. Additionally, a simple user interface may be implemented for user interaction,
allowing functionalities such as initiating face recognition, viewing attendance, or adding
new individuals to the database.
13
8. References
• Student Attendance System using Face Recognition (2020) - Samridhi Dev, Tushar
Patnaikb
https://www.studocu.com/in/document/adikavi-nannaya-university/computer-science-
and-engineering/dev2020-check-this-outt/46627799
• FAREC - CNN Based Face Recognition (2016) - Sharma S, Karthikeyan
Shanmugasundaram, Sathees Kumar Ramasamy
https://www.researchgate.net/publication/312963881_FAREC_-
_CNN_based_efficient_face_recognition_technique_using_Dlib
• Haar Cascade Classifiers in OpenCV
https://medium.com/swlh/haar-cascade-classifiers-in-opencv-explained-visually-
f608086fc42c
• KNeighborsClassifier:
https://scikitlearn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.
html
• OpenCv :
https://docs.opencv.org/4.x/
14
