A web-based facial recognition attendance system
Yadav, Umesh (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025100125450
https://urn.fi/URN:NBN:fi:amk-2025100125450
Tiivistelmä
The increasing demand for secure and efficient attendance management systems has highlighted the necessity of intelligent solutions that reduce human error and limit dependence on external hardware devices.
This research addresses the shortcomings of conventional attendance tracking approaches, which are often unreliable, and constrained by the use of physical devices such as Radio-Frequency Identification (RFID) cards or biometric scanners.
To mitigate these limitations, an Artificial Intelligence (AI)-driven attendance and meeting management system was designed and implemented. The system was developed using Python, Flask (a lightweight web framework), PostgreSQL (a relational Database Management System), and Dlib (a machine learning toolkit). Facial embedding’s were employed to enable real-time identity verification, thereby eliminating manual input and peripheral hardware requirements. Core functionalities include webcam-based authentication, role-based access control, meeting scheduling, automated attendance tracking, and report generation in Comma-Separated Values (CSV) format. Object Relational Mapping (ORM) was achieved through SQLAlchemy, while a responsive and interactive User Interface (UI) was developed using HTML, Bootstrap, and JavaScript. System security was reinforced through password hashing, secure file handling, and robust session management mechanisms.
Experimental evaluation demonstrated an accuracy of 95.6% under optimal lighting conditions. Nevertheless, system performance deteriorated in low-light environments, and mobile compatibility was not incorporated in this version.
Despite these limitations, the study establishes a reliable foundation for future enhancements, including mobile platform integration and advanced analytical features, thereby contributing to the broader development of intelligent, AI-enabled attendance and meeting management systems.
This research addresses the shortcomings of conventional attendance tracking approaches, which are often unreliable, and constrained by the use of physical devices such as Radio-Frequency Identification (RFID) cards or biometric scanners.
To mitigate these limitations, an Artificial Intelligence (AI)-driven attendance and meeting management system was designed and implemented. The system was developed using Python, Flask (a lightweight web framework), PostgreSQL (a relational Database Management System), and Dlib (a machine learning toolkit). Facial embedding’s were employed to enable real-time identity verification, thereby eliminating manual input and peripheral hardware requirements. Core functionalities include webcam-based authentication, role-based access control, meeting scheduling, automated attendance tracking, and report generation in Comma-Separated Values (CSV) format. Object Relational Mapping (ORM) was achieved through SQLAlchemy, while a responsive and interactive User Interface (UI) was developed using HTML, Bootstrap, and JavaScript. System security was reinforced through password hashing, secure file handling, and robust session management mechanisms.
Experimental evaluation demonstrated an accuracy of 95.6% under optimal lighting conditions. Nevertheless, system performance deteriorated in low-light environments, and mobile compatibility was not incorporated in this version.
Despite these limitations, the study establishes a reliable foundation for future enhancements, including mobile platform integration and advanced analytical features, thereby contributing to the broader development of intelligent, AI-enabled attendance and meeting management systems.