Palm Print Recognition Based on Sub-Block Energy Feature Extracted by Real 2D-Gabor Transform

2D-Gabor transforms are considered as an effective spatial-frequency analysis technique in diverse area of image processing, especially in texture feature detection field due to it owns good localization ability in both spatial and frequency domain and also has excellent directional selectivity. In this paper, a method of feature extraction of palm print using real-Gabor transform (RGT) is proposed, which converts the spatial domain information of palm print to joint spatial-frequency domain. In critical sampling case, by calculating the compactly distributed coefficients of RGT, the sub-block energy distribution of palm print in spatial-frequency domain are extracted as recognition features. Experimental results show that this kind of feature has satisfactory discrimination. The proposed feature extraction method has low computational complexity and is highly suitable for palm print recognition due to the time-saving operation. It can achieve high verification accuracy and has favorable robustness against small-scale changes and angle rotation when using different sampling intervals.