Linear Feature Extraction And Its Application To Face Recognition

This paper studied the linear feature extraction in face recognition application. First of all, the paper outlines the background of face recognition and face recognition technology, gives the modular design framework of face recognition system , and introduces a variety of feature extraction methods and characteristics of the application.This paper focus on the main method about linear feature extracting in face recognition recently, that is the PCA and the LDA. Firstly, present the K-L transform and Fisher's criterion in details. Secondly,it focus on the PCA and the LDA of linear feature extracting methods, and present their application in face recognition. Finally, it designs a simple face recognition system to prove the validity, and the experiment results show that the algorithms are accessible, and the systems is also capable to some extent.This paper gives the comparison between the PCA and the LDA about the experiment results. It shows that the LDA is superior to the PCA in the hand of recognition rate. Although both the PCA and the LDA can reduce the demonsions of the feature space , the feature which the PCA attract is optimal set of description feature, not the optimal set of discriminant feature.Paper also gives a simple description about the face recognition trend.