IMPLEMENTATION OF AN IDENTIFICATION SYSTEM WITH FACIAL IMAGE PROCESSING (EIGENFACE) USING MATLAB APPLICATION

Abstract

Facial recognition has emerged as a prominent personal identification system, especially in the industrial sector where traditional attendance card machines have been prevalent. However, manual systems pose several drawbacks, including susceptibility to fraud, lack of flexibility, and resource wastage in card production. To address these issues, this research proposes a shift from card-based attendance systems to a facial recognition-based system using MATLAB application. The study aims to design an identification system based on the eigenface method to process facial images. It begins with a thorough literature review to gather relevant references. Subsequently, a specialized MATLAB application was developed for face identification, and its accuracy was tested. The research utilizes trained data consisting of 45 photos and tested data consisting of 15 photos to evaluate the system's accuracy. The test results reveal a 100% accuracy rate in system identification. Notably, the accuracy varies when identifying faces with different background images, indicating the robustness of the MATLAB application. Overall, the findings suggest that MATLAB can effectively implement an image processing attendance system using the eigenface method.