A new rearrange modular two-dimensional LDA for face recognition
Huxidan Jumahong, Wanquan Liu, Chong Lu · 2011
In this paper, we propose a novel Rearranged Modular 2DLDA(Rm2DLDA) algorithm for face recognition, In the proposed algorithm, the original images are first divided into modular blocks. Then the sub-images are rearranged to form a two dimensional matrix. Two scatter matrices are constructed directly using all the arranged matrices and eigenvectors are derived for image feature extraction based on two dimensional linear discriminant analysis (2DLDA). Experimental results on ORL, YaleB and PIE show that the proposed method can obtain better recognition accuracy.