A palmprint recognition method based on locality preserving projection(LPP) and kernel direct discriminant analysis(KDDA) was proposed in order to improve the recognition performance.In small-size-sample cases(such as image recognition),in order to resolve the singularity of the eigenvalue equation,the image down-sample was first applied to reduce the palmprint space dimension,the LPP was used to extract the local structure features as the input classification extraction features of KDDA,and the cosine dist...