Research on 2D Face Representation and Recognition

Abstract

Considering the effect of the face pose, expression and illumination, a novel feature selection scheme and a face recognition method based on sparse representation were proposed in this paper, which focused on face feature representation and classification. First, the principal component analysis (PCA) was applied on the original samples to remove redundancies and promote the calculate efficiency. After that, a feature selection scheme based on the linear discriminant analysis (LDA) was designed to extract the face features which were more suitable for classification. Finally, the sparse representation framework (RS) was adopted to collect all the face features, and address the recognition task. The experiments tested on FERET, YALE and ORL face database, demonstrate the proposed method is effective and robust to facial pose, expression and illumination conditions to some extent.