Finger Vein Recognition Based on a Histogram of Competitive Gabor Directional Binary Statistics

Traditional methods of extracting finger vein texture changes and orientation features are susceptible to illumination, translation, noise, rotation, the process has difficulty directly structural through original image. In this paper, histogram competitive Gabor directional binary statistics (HCGDBS) is proposed extract discriminative features. First, index largest filter value obtained based on multidirectional as dominant direction, thereby obtaining rotation-invariance feature. Second, according response size each pixel in different directions, order difference relationship between adjacent three directions compared, a highly direction pattern (CGDBP) constructed. Finally, CGDBP extracted blocks, HCGDBS constructed overcome image translation. Experimental results show that it improves recognition performance overcomes rotation.