We know that Principal Components Analysis(PCA) can represent each face image in terms of a linear combination of the eigenface

we also know that the PCA algorithm gives the best representation of images under the sense of minimum mean square error. However

PCA only compares the Euclidean distance between projection coefficients of samples and ignores the residue between the original sample and its reconstructed one. Therefore a new concept called dissimilarity distance metric is proposed in this paper. We project the two images into the same subspace and then characterize the similarity between pairs of samples by comparing to both the projecting coefficients and the approximation errors simultaneously. The higher is the value

the more dissimilar are the two samples. Different from Locality Preserving Projections

a new method

called Dissimilarity Preserving Projections

uses the concept of the dissimilarity above

and constructs the dissimilarity scatter matrix. This algorithm does not have to pre-set the number of neighbors

finally it gets the optimal projection subspace by maximizing the Objective function. The experimental results on AR and FERET face image database demonstrate the effectiveness of the proposed method.