Kernel-based 2DPCA for Face Recognition

Recently, in the field of face recognition, two-dimensional principal component analysis (2DPCA) has been proposed in which image covariance matrices can be constructed directly using original image matrix. In contrast to the covariance matrix of traditional PCA, the size of the image covariance matrix using 2DPCA is much smaller. As a result, it is easier to evaluate the covariance matrix accurately, computation cost is reduced and the performance is also improved. In an effort to improve and perfect the performance efface recognition system, in this paper, we propose a Kernel-based 2DPCA (K2DPCA) method which can extract nonlinear principal components based directly on input image matrices. Similar to Kernel PCA, K2DPCA can extract nonlinear features efficiently instead of carrying out the nonlinear mapping explicitly. Experiment results show that our method achieves better performance in comparison with the other approaches.face r