FACE RECOGNITION ATTENDANCE MANAGEMENT SYSTEM

In today’s fast-paced and technologically advancing world, educational institutions and professional environments are increasingly dependent on digital systems for maintaining efficiency and accountability. One such area that demands automation and reliability is the attendance management system. Traditional methods of attendance, which include manual register signing, roll-calling, or biometric-based systems such as fingerprint scanners, are either time-consuming, susceptible to manipulation, or require physical contact—posing hygiene concerns, especially in the post-pandemic era. This research introduces an innovative and comprehensive web-based Face Recognition Attendance Management System (FRAMS) that employs the capabilities of artificial intelligence and computer vision to offer a secure, efficient, and non-intrusive solution for recording attendance. The system is developed using a combination of Python, Django, and front-end technologies such as HTML, CSS, JavaScript, and Bootstrap for a fully responsive and interactive user interface. The face recognition functionality is powered by Dlib and OpenCV, two of the most robust and widely-used libraries in the computer vision domain. Dlib is used for extracting high-dimensional facial embeddings while OpenCV handles real-time camera capture and image pre-processing. The system utilizes MySQLite during local development and is scalable to MySQL for production-level database operations. FRAMS introduces a dual-role architecture where both students and teachers can register, log in, and access role-specific dashboards. During student registration, users are required to submit essential academic and personal details such as username, email, password, USN (University Serial Number), department, and a facial image either through direct webcam capture or by uploading a pre-existing photo. This image is then processed, and the encoded facial features are stored in the database. Teachers, during registration, provide their name, email, password, designation, and department. Once authenticated, teachers can initiate the attendance process from their dashboards. This action activates the webcam which captures live footage of students present in front of the system. These faces are then compared in real-time with the database of pre- registered student faces. If a match is found within the set threshold, the system automatically marks the student as “Present” in the attendance log; otherwise, their attendance status remains “Absent.” Students have access to their attendance reports and can view their attendance percentage through their personalized dashboard. The primary objectives of this system are to eliminate human error, reduce administrative workload, and enhance accuracy and integrity in attendance tracking. The facial recognition component is trained to achieve a high level of accuracy under varied lighting conditions and angles to ensure reliability. Further, the web application is designed to support scalability, security (with hashed passwords and role-based access control), and performance optimization for real-time use. This paper discusses the architectural design, technology stack, data workflow, recognition pipeline, database schema, and real-time implementation of the system in detail. Additionally, experimental results, accuracy analysis, and user experience metrics are presented to demonstrate the feasibility and effectiveness of the proposed solution in real-world environments..