Research on algorithms of face recognition based on kernel function

The key technique of face recognition and the main characteristics of the kernel function are presented in this paper. Combined with the relative research results that are proposed lately, three algorithms of face recognition based-on kernel function which include kernel principal component analysis (KPCA), kernel fisher discriminating analysis (KFDA) and support vector machine (SVM) are presented. What is more, an experiment is put forward to show that when adopting the kernel method in the process of face recognition there exists technological advantage in the area of the efficiency of feature extraction and the generalization ability of classification.