Palm Vein Recognition by Combining Curvelet Transform and Gabor Filter

Biometrics research based on palm vein recognition has been developed rapidly in recently years. However, due to the poor palm vein image quality, the performance of the recognition is not good enough. Recently, coding algorithms, such as Curvelet transform and Gabor Filter have been proposed and have been attracting much research attention. While the Curvelet Transform is good at extracting the linear features from the palm vein images, the Gabor Filter excels in extracting the orientation features. By investigating these two different coding schemes, we propose in this paper a score-level fusion scheme for palm print/vein verification. The proposed method was applied on the HK PolyU Database and an EER of 0.1023% was achieved, which outperforms using the Curvelet Transform or Gabor Filter alone.