A Palmprint Recognition Method Based on Gabor Wavelet and Local Direction Pattern

摘要

Palmprint recognition is an emerging biometric technology that identifies a person based on the hand image. In this paper, a novel palmprint identification method based on Gabor wavelets and Local Direction Pattern (LDP) techniques has been proposed. The Gabor response images at different scales and directions are obtained by convolving the multi-channel Gabor wavelet and palmprint images, which are further divided into non-overlapping sub-blocks. The LDP characteristics are extracted from each sub-block and the feature vector of palmprint image is then produced by combining them using the LDP operator. Finally, the palmprint images are identified using the nearest neighbor classification method. The proposed method has been verified on the public PolyU palmprint database, and the experimental results have shown that the proposed method is effective.