Synergy in Facial Recognition Extraction Methods and Recognition Algorithms

Abstract

This paper aims to survey on the existing research works done on facial recognition and acknowledge their differences. Understanding facial recognition processes such as facial normalization, facial detection, facial extraction and facial recognition methods and algorithms are part of the essence of this paper. This paper outlines the purposes of existing techniques as well as its challenges. This paper also looks at the idea whether combining several techniques is feasible in order to produce a better synergy result. Methods are evaluated based on their classification rate percentages as well as the numbers of dimensionality reductions. Based on the literature reviews conducted, the facial recognition algorithm is made up of two steps. The first step is when an individual model is modeled in the database based on the color appearance and geometrical information provided by the available images whereby each model characterizes an individual like a bar code or a unique serial number and discriminates it from the other people in the database. The second step is to carry out the identification using a classifier, related with the standard Gaussian distribution, to decide whether a face image belongs to one person in the database or not. This paper has performed a comparative analysis of previously conducted experiments and based on the findings obtained, a schema of the framework for face recognition is proposed.