2019 IEEE 9th Annual Computing and Communication Workshop and Conference (CCWC) | 2019
Design and Implementation of a Hybrid Face Recognition Technique
Abstract
The problem of face recognition is a subject that has been researched for many years. Face recognition technology can be utilized in many different industries, it isn’t solely designated for computer science relate fields. The purpose of this paper is to continue this research and attempt to provide a powerful classifier that works on par with the algorithms currently on the market. The proposed classification algorithm in this paper will be known as the EFL Hybrid algorithm and it is based on combining three face recognition algorithms to provide a higher accuracy. This classifier is created by combining three major algorithms that will be explored in this paper: Eigenfaces, Fisherfaces, and Local Binary Patterns Histograms. The experimental results show that both the LBPH and Fisherfaces classifiers perform better than the Eigenfaces classifier. The experimental results on the testing datasets show that the hybrid face recognition technique can provide a higher accuracy compared to other face recognition algorithms.