Increasing the distinctiveness of hand vein images by oriented gradient maps

Abstract

Hand vein pattern as a biometric trait for people identification has attracted an increasing interest in the recent years thanks to its properties of uniqueness, permanence, non-invasiveness and strong immunity to forgery. In this paper, we propose to make use of Oriented Gradient Maps (OGMs), that we previously proposed for face recognition using the term of Perceived Facial Images (PFIs), to represent Near-Infrared (NIR) hand dorsa vein images and to highlight the distinctiveness of hand vein patterns. Using a holistic approach through both the popular PCA and LDA, we benchmarked the proposed hand vein representations on the NCUT dataset of 2040 hand vein images and demonstrated the effectiveness of the proposed hand vein representations.