Improved-LDA based face recognition using both facial global and local information

摘要

To achieving higher classification rate under various conditions is challenging task in face recognition community. This paper presents a combined feature Fisher classifier (CF~2C) approach for face recognition, which is robust to moderate changes of illumination, pose and facial expression. The novelty of the method are: (1) the facial combined feature used for face representation, which is derived from facial global and local information extracted by DCT and (2) the development of Fisher classifier for high-dimensional multi-classes problem. Experiments on ORL and Yale face databases show that the proposed approach is superior to the traditional methods such as Eigenfaces and Fisherfaces.