A new PCA based approach, eigenblock, for face recognition is proposed in this paper. Face images are first partitioned into face blocks and then PCA is performed based on the blocks. To cope with the problem of local occlusion and distortion of face images, an improved minimum distance classifier which only counts a number of best-matched blocks is also presented. Theoretical analysis and experimental results show that the proposed approach is superior to the conventional PCA approach in recognition speed, real time algorithm realization and robustness to occlusion.
Face recognition, PCA, Eigenface, Eigenblock
