Biometric Authentication by Finger Vein and Knuckle using LBP and Gabor

Biometric technologies are automated methods of verifying or recognizing the identity of a living person based on a physiological or behavioral characteristic. Biometrics technology is without doubt a more efficient way of authentication than the more common use of passwords, smart cards, or a combination of the two. Passwords also have expiration dates that require new assignment of passwords and also users cannot remember their pass words, trying to navigate through a series of steps to access needed information, becomes time consuming. Biometrics is a promising technology that remains as a solution to these problems.The ideal biometric characteristic has five qualities namely robustness, distinctiveness, availability, accessibility and acceptability. Finger knuckle and vein satisfies these as they are unique and many samples are available per hand and independent to any behavioral aspect. They are user-centric, contactless and unrestricted access control. In this paper, both the major and minor knuckle and finger vein are used so as to improve the accuracy. Here, the input images are preprocessed and ROI segmented for better results. The knuckle and vein image features are then extracted using the Local Binary Pattern and Log Gabor techniques separately for knuckle and vein and the resultant features are combined for classification using the K-Nearest Neighbor as classifiers. The results are analyzed and this proposed method yields 94% accuracy and the performance is also increased. Keywords—finger knuckle, finger vein, local binary pattern, gabor, k-nearest Neighbor