Conceptual Model of the Smart Attendance Monitoring System Using Computer Vision

The monitoring of student attendance remains a fundamental and vital part of any educational institution. Although some automated biometric-based attendance monitoring systems have been proposed in the literature, poor system modeling has negatively affected their performance. There is, hence, a need to address the various needs associated with the development of an AI-based attendance monitoring system. This paper proposes the conceptual model for a smart attendance monitoring system that uses face recognition to monitor students’ attendance during lectures. The front and back-end system architectures, as well as a full view multi-camera structure aimed at effectively capturing and detecting of faces, are presented. Thee different feature extraction techniques - PCA, LDA, LBP and a combination of PCA and LDA were investigated. The model achieved a recognition accuracy of 90%.