An Extensive Survey on Feature Extraction Techniques for Facial Image Processing

Abstract

In this research paper an extensive literature survey on different types of feature extraction techniques is reported. To provide an extensive survey, we not only categorize existing feature extraction techniques but also provide detailed descriptions of representative approaches within each category. These techniques are simply classified into four major categories, namely, feature based approach, appearance based approach, template-based and part-based approaches. The aim of this paper is to report an illustrative and comparative study of most popular feature extraction methods which are generally used in face recognition problems. This paper provides an up-to-date comprehensive survey of existing face recognition researches. We are motivated by the lack of direct and detailed independent comparisons of all possible algorithm implementations in available literature. After extensive research on these feature extraction techniques we found that different feature extraction techniques yield prominent results for different image processing applications.