FACIAL FEATURE BASED STUDENT ATTENDANCE AUTOMATION USING DEEP LEARNING

Abstract

Conventional attendance recording methods in academic institutions — manual roll calls and paper registers — are inefficient, time-consuming, and prone to errors including proxy attendance. This paper presents a Smart Attendance System that leverages real-time face recognition to automate the entire attendance workflow. The proposed system employs the Local Binary Pattern Histograms (LBPH) algorithm for face detection and recognition, integrated with a Flask-based web application, a MySQL relational database, and an automated email notification module. The system supports a four-punch mechanism (Early In, Late In, Lunch Out, Lunch In, and Day Out) to record granular daily attendance. An academic marks management module records mid-term examination results and dispatches grade reports via email. Experimental evaluation demonstrates 94.1% recognition accuracy under realistic lighting conditions with a false acceptance rate of 1.2%. Role-specific dashboards for administrators and students enable real-time monitoring, report generation, and condonation tracking. The results confirm that the system significantly reduces administrative overhead while improving data accuracy and institutional transparency.