EVALUATING THE EFFECTIVENESS AND PERFORMANCE OF AN EXAMINATION HALL ATTENDANCE SYSTEM WITH HIGH-PERFORMANCE FACE RECOGNITION AND FINGERPRINT TECHNOLOGY

Abstract

Attendance systems are vital in educational settings, particularly higher institutions, for ensuring academic integrity and efficient administration. Traditional methods, such as manual roll calls, are prone to human error and fraud. This paper investigates the effectiveness and performance of an examination hall attendance system that integrates high-performance face recognition and fingerprint technology. The proposed system utilizes high-resolution cameras, biometric sensors (fingerprint and voice), a memory card module, and an RGB digital display connected to a Raspberry Pi. Data is processed using machine learning algorithms, including Convolutional Neural Networks (CNNs) for face recognition and Python-based biometric authentication software. The system features a computer interface with memory card storage for monitoring and control, enhancing accuracy, efficiency, and security in exam administration. Field tests demonstrate that the system effectively discourages attendance manipulation and achieves accurate student identification under challenging conditions. The results show an accuracy of 98.8%, a false acceptance rate (FAR) of 0.6%, and a false rejection rate (FRR) of 1.0%. The system processes attendance in 1.2 seconds, significantly improving efficiency over manual methods.