论文信息 - Finger vein identification based on 2-D Gabor filter

Finger vein identification based on 2-D Gabor filter

In order to improve the efficiency of finger vein identification, this paper proposed a new identification method based on the direction feature of the vein image. Firstly, 2-D Gabor filter is used to filter the original vein image. Then the phase and direction texture features are extracted. An algorithm is used for fusion at feature level. Finally, a modified Hamming distance measure is used for matching. The proposed method does not require any image pre-processing and positioning. It is easy to use and has a good performance on extracting the vein feature with an identification rate of 99.31%.

Wang Kejun | Liu Jingyu | Feng Weixing | P Popoola Oluwatoyin

[1] David Zhang,et al. Fusion of phase and orientation information for palmprint authentication , 2005, IEEE International Conference on Image Processing 2005.

[2] John Daugman. How iris recognition works , 2004 .

[3] Yoshihiko Hamamoto,et al. Normalization techniques of handwritten numerals for Gabor filters , 1997, Proceedings of the Fourth International Conference on Document Analysis and Recognition.

[4] Wang Ke. Finger Vein Recognition Based on Wavelet Moment Fused with PCA Transform , 2007 .

[5] John G. Daugman,et al. Complete discrete 2-D Gabor transforms by neural networks for image analysis and compression , 1988, IEEE Trans. Acoust. Speech Signal Process..

[6] Naoto Miura,et al. Feature Extraction of Finger-vein Patterns Based on Repeated Line Tracking and Its Application to Personal Identification , 2022 .

[7] David Zhang,et al. Palmprint identification using feature-level fusion , 2006, Pattern Recognit..

[8] John Daugman,et al. High Confidence Visual Recognition of Persons by a Test of Statistical Independence , 1993, IEEE Trans. Pattern Anal. Mach. Intell..