Human Identification Based on Hand Dorsal Vein Pattern using BRISK & SURF Algorithm

Abstract

Security is an essential part of our lives today as people have many accounts and each account generally has a high amount of transactions. Biometrics are in the trend as it reduces risk and is quite difficult to replicate. Currently most popular biometrics such as fingerprint, retina can be forged. Dorsal Vein pattern is gaining attention these days due to its contactless nature and difficult to forge. These vein patterns can give rise to another reliable biometric which is better than the biometrics being currently used. In this paper, we have discussed the technique to extract the important information in the form of points and use them for analysis based on different parameters. The analysis will help us to know the reliability of biometrics using hand dorsal vein pattern. We have used SUAS Dorsal Hand Vein Database which is available online [18]. We have implemented the technique on MATLAB - R2019 image processing toolbox. We have used end points and branch points of the vein pattern to form triplets using Delaunay’s Triangulation. Then these triplets are matched using proposed matching technique. We have used matching techniques in MATLAB for analysis. We have used the BRISK and SURF algorithm for matching. The objective of this paper is to build a distinct and a substitute to the contact-based systems. Vein pattern biometrics are quite useful in reducing crimes as it is a contactless biometric.