Design and application of Compound Kernel-PCA algorithm in face recognition
Chengyuan Liu, Ting Zhang, Dong-Sheng Ding, Chongshan Lv · 2016
The principal component analysis (PCA) is one of the most commonly used feature extraction methods in face recognition, but the traditional PCA method can't deal with the non-linear problem between pixels. In this paper, based on the traditional PCA, combined with the advantages of KPCA(kernel-PCA), a new composite kernel-PCA algorithm is designed. By combining the two single kernel functions, the new algorithm can make full use of their complementary characteristics. Experiments were performed on ORL and FERET face database respectively. Through the analysis and comparison of the experimental results, it is proved that this algorithm can achieve the efficient recognition of face images, and has better robust performance when dealing with large sample database.