Palmprint recognition using contourlets-based local fractal dimensions

A novel efficient palmprint recognition method called contourlets-based local fractal dimensions (CLFD) is represented in this paper. The novelty of CLFD comes from using fractal dimension independent of fractal coding for palmprint image representation. Three main steps are involved in the proposed CLFD: (i) Contourlets subbands are extracted by the convolution of contourlet bank and the original gray images; (ii) For more local features, all the contourlet subbands are partitioned into small uniform blocks whose fractal dimensions are computed to form the feature vectors; and (iii) The Manhattan distance and the nearest neighbor classifier are finally used for classification. The method is not only robust to the variations and distortions occurred on palmprint images, but also efficient in feature extraction and matching. The effectiveness of the proposed method is demonstrated by the experimental results.