Finger Vein Pattern Extraction Improvement by Enhance Maximum Curvature and Frangi Filter
Việt Dũng Nguyễn, Anh Tu Tran · 2022 13th International Conference on Information and Communication Technology Convergence (ICTC) · 2022
The usage of fingerprint pattern for identification is very common. Finger vein pattern is known less than fingerprint pattern, but finger vein pattern has a better security aspect. Finger vein pattern is already commercially exploited but the scientific research done is still lacking behind. Finger vein is unique to each individual and it only be observed by infrared light. Recently, the identification based on finger vein feature is the trends of biometric identification system because finger vein recognition method has been proved to be a convenient, effective and secure. Because of vein pattern from low quality vein images, feature extraction is to be the most important step in the vein recognition process. In this work, a method to enhance extraction of finger vein pattern by using enhance maximum curvature and Frangi filter is proposed. The experimental results in The University of Twente Finger Vascular Pattern (UTFVP) database show that the proposed approach efficiently extracts the finger vein texture feature, therefore improves recognition performance.