Hand vein authentication system using connected minutiae neighbours

Abstract

This paper proposes an efficient personal authentication technique using hand vein patterns. A dynamic region of interest (ROI) extraction algorithm is presented through which more features can be extracted compared to the fixed ROI. The proposed system extracts the minutiae (bifurcation and end points) of the vein pattern after a series of preprocessing steps such as image enhancement, noise removal, binarisation and thinning. The system then computes the connected minutiae neighbours of each minutiae in the vein pattern. Finally, a matching algorithm that uses the minutiae neighbour information is applied for authentication. The proposed system is evaluated using a vein pattern database of 194 images. It has been found that the proposed algorithm has performed better compared to others and has been able to achieve an equal error rate of 1.2%.