Application of Kernel Method on Face Feature Extraction

Principle component analysis (PCA) and linear discriminant analysis(LDA) are effective methods for feature extraction, and they were successfully utilized for face recognition. But face image data distribution in practice is highly complex because of illumination, facial expression and pose variations. So it is necessary to extract nonlinear features for face recognition. Both PCA and LDA are performed by only using the second-order statistics among image pixels, and not sensitive to high order statistics in the data. In this paper, the kernel trick is used firstly to project the original samples into an implicit space called feature space by nonlinear kernel mapping, then the high order relations are extracted in high dimensional space. The experiment shows that the nonlinear feature extraction methods are effective and outperform the traditional methods.