Possilble Applications of Face Recognition Software in Enhancing School Security Systems

Abstract

In this paper, we propose using the face recognition software developed by the Faculty of Mathematics of the University of Belgrade to enhance school security systems. We give an overview of biometric identification methods, with special focus on automated face recognition software. We present the basic steps of the face recognition algorithm taken by the software and the existing methods from literature that could be employed in each individual step. We then describe the face recognition software developed by the research team at the Faculty of Mathematics. The software employs the concept of hybridization of several methods in every step of the algorithm to maximize the advantages of each method. We present the excellent evaluation results of the proposed software, obtained using known testing databases from literature. We then propose two models for employing the face recognition software to identify individuals and control their access to school premises. We discuss and analyze both models and focus on the process of setting the parameters for each model in order to obtain better results.