Competitive Coding Scheme based on 2D Log-Gabor filter for Palm Vein Recognition

摘要

This paper proposes a novel palm vein recognition method based on a competitive coding scheme using 2D log Gabor filtres. The proposed method consists of two major steps: (i) inspired by the bitwise competitive coding, the feature extraction employs 2D log Gabor filtering where the final feature map is composed by the winning codes of the lowest filters' bank response, and (ii) the matching process uses the Jaccard distance as a metric to capture efficiently the similarities between the feature maps and allowing to make a decision. The experiments carried out on the challenging MS-Polyu database have shown that the proposed method yields a significant performance gains compared to existing state-of-the-art methods.