Two modality-based bi-finger vein verification system
Zhichao Li, Sun Dongmei, Di Liu, Hao Liu · 2010
At present, the finger vein has attracted more attention, but the existing finger vein extraction systems are single finger systems, multi-finger case is not yet to come, and few work was did about the research of corresponding verification algorithms. This paper proposes a two modality-based bi-finger vein verification system. Both the finger vein and finger shape could be extracted from the single image acquired from the sensor. The system includes the new finger vein network extraction algorithm. The intersection of the forefinger and middle finger as the origin is introduced to the coordinate system, proposing a method including determining ROI, the finger vein and shape features extraction and its corresponding fusion verification. The experiment shows that the new finger vein network extraction algorithm can extract the finger vein network well. The introduction of the coordinate system effectively solves the problem that not easy to determine ROI, and the false accept rate of bi-pattern fusion verification system is much smaller than the false accept rate of single-feature fusion verification system.