An improved null space LDA method based on discrete cosine transform (DCT) is proposed to solve both problems. First, by employing DCT instead of "pixel grouping" process and using different definition of the within-class scatter matrix, an improved null space method is proposed. Then, we combine this method with fractional-LDA (F-LDA) and an efficient new feature extraction algorithm to perform face recognition. Experimental results show that the proposed method achieves better performance than other existing ones.
展开▼