Kathmandu University
Department of Computer Science and Engineering
Dhulikhel, Kavre
A Project Report
on
“Smart-Attend”
[Code No COMP 206]
(For partial fulfillment of Third Year/ First Semester in Computer Science)
Submitted by
Binayak Parajuli (45)
Kapil Adhikari(47)
Ashwin Sapkota(48)
Submitted to
Mr. Suman Shrestha
Department of Computer Science and Engineering
Submission Date:
01/07/2025
1
Bona fide Certificate
This project work on
“Smart-Attend”
is the bona fide work of
“Innima Karki (19)
Binayak Parajuli (53)”
who carried out the project work under my supervision.
Project Supervisor
________________________
Sameer Tamrakar
Assistant Professor
Department of Computer Science and Engineering
Date:01/07/2025
2
Abstract
In today’s fast-paced world, the need for efficient and accurate attendance tracking
systems has become increasingly important, especially within educational institutions
and corporate environments. Traditional methods of attendance management often
suffered from inaccuracies, inefficiencies, and susceptibility to fraudulent practices.
To address these challenges, we developed a Face Recognition Attendance System
(FRAS) that leverages modern technologies to enhance accuracy, security, and
convenience.
The system utilized state-of-the-art facial recognition technology powered by
machine learning algorithms to identify individuals based on their unique facial
features. This eliminated the need for physical contact or manual data entry. The
solution was implemented using a web-based interface built with ReactJS for the
front-end, Flask for the back-end, and OpenCV for facial recognition tasks. Data
storage and retrieval were handled efficiently using MySQL.
Upon completion, the Face Recognition Attendance System demonstrated high
accuracy and performance during testing. It significantly reduced the time and effort
required for attendance management while improving administrative oversight and
eliminating common issues such as proxy attendance. Overall, the project fulfilled its
objectives and provided a secure, user-friendly, and scalable attendance tracking
solution.
Keywords: Face Recognition, Attendance System, Machine Learning, OpenCV,
ReactJS, Flask
1
Table of Contents
Abstract.........................................................................................................................1
Table of Contents......................................................................................................... 2
Abbreviations............................................................................................................... 3
Chapter 1 Introduction................................................................................................1
1.1 Background........................................................................................................ 1
1.2 Objectives.......................................................................................................... 2
1.3 Motivation and Significance.............................................................................. 2
Chapter 2 Related Works............................................................................................3
2.1 Face Attendance.................................................................................................3
2.2 Clockgogo.......................................................................................................... 4
Chapter 3 Design and Implementation...................................................................... 5
3.1 System Architecture...........................................................................................5
3.3 Facial Recognition Module................................................................................5
3.4 Development Process.........................................................................................6
3.5 ER diagram........................................................................................................ 8
3.6 Data Flow of the system.................................................................................... 9
3.7 SYSTEM REQUIREMENT & SPECIFICATIONS........................................10
3.7.1 Software Requirements...........................................................................10
3.7.2 Hardware Requirements..........................................................................11
Chapter 4 Discussion on the achievements..............................................................12
4.1 Features............................................................................................................ 12
Chapter 5 Conclusion and Recommendation..........................................................13
5.1 Limitations....................................................................................................... 13
5.2 Future Enhancement........................................................................................ 14
References................................................................................................................... 14
APPENDIX.................................................................................................................15
2
Abbreviations
FRAS Face Recognition Attendance System.
GPS Global Positioning System.
OpenCV Open Source Computer Vision Library
3
Chapter 1 Introduction
1.1 Background
Traditional attendance systems, such as manual registers or card-based check-ins, had
long been used in educational and corporate settings. However, these methods often
suffered from various limitations, including susceptibility to human error,
time-consuming processes, and vulnerability to fraudulent practices such as proxy
attendance. As institutions began to demand more reliable and automated solutions,
the limitations of these outdated systems became increasingly evident.
In response to these challenges, we developed a Face Recognition Attendance System
(FRAS) designed to streamline and secure the process of attendance tracking. The
system employed advanced facial recognition algorithms to identify individuals based
on their unique facial features, eliminating the need for physical interaction or manual
data entry.
This technology-driven approach not only improved the accuracy and speed of
attendance collection but also enhanced the overall security and integrity of the
process. The successful development and implementation of the FRAS system
demonstrated how modern computer vision and machine learning techniques could be
effectively applied to solve real-world administrative problems in educational and
organizational contexts.
1
1.2 Objectives
● To accurately recognize individuals' faces and minimize errors commonly
found in traditional attendance systems
● To automate the attendance tracking process, thereby reducing manual effort
and saving valuable time for both users and administrators.
● To develop a user-friendly interface that allowed users to check in and out
seamlessly without requiring physical cards or manual entries.
● To provide real-time access to attendance records for administrators, enabling
them to monitor, manage, and analyze attendance data effectively.
1.3 Motivation and Significance
The motivation behind this project stemmed from the inefficiencies and limitations of
traditional attendance systems such as paper registers and card-swiping mechanisms.
These methods were often prone to human error, time-consuming procedures, and
security concerns like buddy punching or proxy attendance. As institutions evolved
and embraced digital solutions, the need for a more accurate and secure attendance
tracking system became evident.
To address these issues, we implemented a facial recognition-based attendance
system that streamlined attendance management while enhancing reliability and
accountability. The goal was to eliminate the shortcomings of manual systems by
leveraging automated facial recognition, thereby saving time and improving accuracy.
The project offered a modern solution suited for both physical and remote
environments. By ensuring that only authorized individuals could register attendance,
the system improved access control and institutional transparency. Overall, it marked
an innovative step toward digital transformation in attendance management,
improving operational efficiency and user experience.
2
Chapter 2 Related Works
2.1 Face Attendance
The Face Attendance (Face Attendance) system stands out for its real-time face
detection, which accurately detects live faces from photographs. It operates offline,
simplifies management through user-friendly interfaces, and seamlessly integrates
with HR systems. Supporting multiple platforms, including smartphones and
computers, enhances its versatility. This system’s comprehensive features ensure
efficient employee attendance tracking, bolstering security and productivity across
various workplace environments
3
2.2 Clockgogo
Clockgogo (Clockgogo), a cloud-based time attendance management system,
employs patented 4-level verification technology to streamline time tracking and
alleviate HR concerns. It eradicates problems like fake Global Positioning System
(GPS) and buddy punching while minimizing hardware costs and installation
complexities. Featuring user-friendly roster planning and reporting functionalities, it
simplifies work hours and overtime calculations with just a click. Integration with
third-party systems like Workday and Peoplesoft is seamless through API access.
Since its 2016 inception, Clockgogo has processed millions of faces and garnered
widespread adoption among global brands.
4
Chapter 3 Design and Implementation
3.1 System Architecture
The system was developed as a web-based application consisting of three main
layers:
● Front-end: Implemented using ReactJS, the front-end provided a responsive
and intuitive user interface for students, teachers, and administrators. Users
could view attendance records, mark their presence, and manage their profiles
through a streamlined interface.
● Back-end: Built with Flask, the back-end handled API requests, database
interactions, and facial recognition processing. Flask’s lightweight nature
allowed for rapid development and easy integration with the machine learning
model
● Database: MySQL was used to store user data, attendance records, and image
embeddings. The database schema was designed to efficiently handle user
roles, facial features, and timestamps
3.3 Facial Recognition Module
OpenCV was utilized for real-time face detection and recognition. During
registration, each user's facial image was captured, processed, and stored as a unique
vector using facial embedding techniques. During login or check-in, the system
captured a live image and compared it with stored embeddings using a cosine
similarity metric. The recognition module achieved reliable accuracy during testing,
with minimal false positives.
5
3.4 Development Process
The development process was divided into key phases:
1. Requirement Analysis and Research:
We studied existing systems and defined the technical and functional
requirements for SmartAttend.
2. Wireframing and UI/UX Design:
Wireframes were designed for all pages, including login, dashboard,
attendance view, and admin panel. A clean and consistent UI was
implemented using Tailwind CSS.
3. Front-end Development:
Components were built using ReactJS, with routing handled via React Router.
The interface was designed to be responsive across devices.
4. Back-end Development:
RESTful APIs were developed using Flask to handle authentication, face
registration, attendance logging, and data retrieval.
5. Face Recognition Integration:
The OpenCV library was integrated with the Flask backend to handle face
detection and recognition in real-time.
6. Testing and Debugging:
The system was tested with multiple users under varying lighting conditions
to ensure accuracy and reliability.
6
3.5 Use Case Diagram
7
3.5 ER diagram
8
3.6 Data Flow of the system
9
3.7 SYSTEM REQUIREMENT & SPECIFICATIONS
The SmartAttend system was designed to be lightweight, platform-independent, and
easily deployable across standard computing environments. This section outlines the
key software and hardware components that supported the successful design and
implementation of the project.
3.7.1 Software Requirements
The following software tools and frameworks were used during development:
● JavaScript: Enabled interactive and dynamic behavior on the client-side,
forming the foundation of the front-end functionality.
● ReactJS: Utilized to build the front-end interface in a modular,
component-based architecture. It ensured maintainability, reusability, and
responsiveness across devices.
● Flask: Served as the back-end micro web framework. It handled routing, API
requests, authentication, and face recognition processing.
● MySQL: Used as the relational database system to store user profiles,
attendance logs, and encoded facial feature vectors. It offered stability,
structured querying, and efficient data management.
● OpenCV: Powered the facial recognition module. It provided tools for
real-time face detection, image preprocessing, and embedding generation.
10
3.7.2 Hardware Requirements
The system was developed and tested on readily available hardware without the need
for high-end configurations. The minimum hardware specifications included:
● Camera: A basic web camera (720p or higher) for capturing facial images
during registration and attendance sessions.
● Development Machine:
○ Processor: Intel i3 or equivalent (or better)
○ RAM: Minimum 4 GB
○ Storage: Minimum 10 GB free space
○ OS: Windows, macOS, or Linux with Python and [Link]
environments configured
● Browser Compatibility: The system was fully functional on modern web
browsers such as Google Chrome and Mozilla Firefox, with JavaScript
enabled.
11
Chapter 4 Discussion on the achievements
The SmartAttend project successfully achieved its primary objectives by developing a
functional face recognition-based attendance system using OpenCV, Flask, ReactJS,
and MySQL. The system allows for real-time, contactless attendance marking with
high accuracy, offering a significant improvement over traditional methods. A
user-friendly interface was created for both users and administrators, enabling smooth
interaction and efficient management of attendance data. The integration between the
front-end and back-end ensures secure and reliable data processing. The system was
tested under various conditions and performed consistently well, with minimal errors
and quick response times. Overall, SmartAttend meets its goals of enhancing
efficiency, accuracy, and user convenience in attendance tracking.
4.1 Features
● Face Recognition-Based Attendance: Automatically marks attendance in
real-time using facial recognition, minimizing manual work and errors.
● Class Creation and Management: Enables teachers to create and manage
virtual classes, similar to platforms like Google Classroom.
● Notice Posting: Allows administrators and teachers to post important
announcements and updates for students.
● Notes Upload: Facilitates uploading of study materials and lecture notes for
student access.
● Attendance Sheet Download: Provides options to generate and download
attendance records in a structured format.
● User-Friendly Interface: Designed with a clean, intuitive interface for
smooth navigation and ease of use
12
Chapter 5 Conclusion and Recommendation
The SmartAttend project has successfully delivered a modern and efficient solution
for attendance management by integrating facial recognition technology with class
management features. The system not only automates the attendance process but also
enhances classroom interaction through features such as class creation, notice posting,
and resource sharing. Its user-friendly interface and reliable performance make it
suitable for deployment in educational institutions aiming to improve operational
efficiency and reduce manual administrative tasks.
5.1 Limitations
● Face recognition accuracy depends on lighting and camera quality.
● No mobile app; limited to web access only.
● Basic class management without assignment or grading features.
● No real-time push notifications for notices or updates.
● Lacks advanced security features like role-based access.
● Manual face registration is required for each user.
5.2 Future Enhancement
● Add real-time chat and video conferencing features
13
● Implement automated grading and feedback system
● Integrate with third-party calendar and reminder apps
● Enable multi-language support for wider accessibility
● Develop mobile app versions for iOS and Android
● Incorporate AI-based attendance anomaly detection
References
14
APPENDIX
15
16
17
18
19
20
