Facial Recognition Attendance Monitoring System

Abstract

Attendance tracking is a crucial part of managing academic institutions, but traditional methods—like manual roll calls or card systems—often fall short. They can be inefficient, take up too much time, and are prone to errors and even cheating, such as students marking attendance for their friends. On the other hand, existing biometric solutions that rely on contact, like fingerprint or iris scanning, come with their own set of issues, including hygiene concerns, high hardware costs, and slow processing times for large groups of students. This paper introduces an innovative, contactless Facial Recognition Attendance System aimed at overcoming these challenges in a college setting. The system is designed as a mobile-first web application, allowing students to easily check in using their own smartphones. It utilizes open-source computer vision libraries, particularly [DeepFace] in combination with [OpenCV], to ensure accurate and efficient facial recognition. Additionally, the system makes use of browser-based Geolocation APIs and a backend geofencing module to enforce location-based attendance, confirming that students are actually on campus. A working prototype has been developed and tested on a small scale, showing promising results in real-time facial recognition and location verification through a standard mobile web browser. This project presents a practical and budget-friendly attendance solution that could greatly improve security, cut down on proxy attendance, and boost administrative efficiency in educational environments.

Introduction

The paper presents a novel, contactless Facial Recognition Attendance System designed to improve attendance tracking in colleges. Traditional methods like manual roll calls and card systems are inefficient and prone to errors or cheating, while existing biometric solutions such as fingerprint or iris scans face hygiene, cost, and speed issues.

This system leverages students’ smartphones via a mobile-first web application, using open-source computer vision libraries (DeepFace and OpenCV) for accurate facial recognition and browser-based Geolocation APIs combined with backend geofencing to verify that students are physically on campus. The system operates on a client-server model, with a React-based frontend capturing facial images and location, and a Python Flask backend handling face recognition, geofencing, and data management.

A prototype was developed and tested, demonstrating effective student enrollment and attendance marking, rejecting proxy attendance and verifying location within campus boundaries. The system is cost-effective, requiring no specialized hardware, and accessible on standard smartphones without native app installation.

However, the prototype has limitations: it lacks liveness detection (making it vulnerable to spoofing), uses in-memory storage (losing data on server restart), relies on a simplistic geofencing method, and is not optimized for large-scale deployment or comprehensive administrative features.

Conclusion

In conclusion, this project successfully designed and implemented a functional prototype of a mobile-based facial recognition attendance system incorporating geofencing for college students. By leveraging open-source libraries such as [DeepFace] and [OpenCV] for facial authentication, alongside standard web technologies for mobile accessibility and geolocation, the system offers a contactless and potentially more secure alternative to traditional attendance methods. The prototype demonstrates the core capabilities of student enrollment based on facial embeddings and location-validated attendance marking via a smartphone web browser. While the current implementation has limitations regarding liveness detection, database persistence, and geofencing precision, it successfully serves as a proof-of-concept, validating the core idea and highlighting the potential for such systems to improve efficiency and reduce proxy attendance in educational environments using readily available technology.

References

[1] Documentation for DeepFace, OpenCV, Flask, React, Vite, Axios, React Router, Tailwind CSS, Python, Flask-CORS, python-dotenv. [2] Academic papers related to facial recognition algorithms (like Facenet, ArcFace if used by DeepFace\'s default model). [3] Academic papers or sources on attendance systems, geofencing, or biometric security in educational settings. [4] IPinfo.io. “Free IP Geolocation API.” https://ipinfo.io [5] Python Software Foundation. (2024). Python 3 Documentation. https://docs.python.org [6] Flask Documentation. (2024). https://flask.palletsprojects.com

Copyright

Copyright © 2025 A. Rajamurugana, Rithickroshan S, Saneesh Kumar V, Mathanraji V. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.