IntroductionThe rapid growth in the use of e-commerce applications requires reliable user identification for effective and secure access control. The face identification beyond visible spectrum is increasing receiving attention and has highest user acceptance [23], [7]. Hand vein identification has emerged as a promising component of biometrics study.The subcutaneous vascular pattern/network appearing on the back of hand, referred to as the hand vein in this paper, is extremely difficult to forge and therefore offers promising These changes in vascular system make the vein pattern loose and change its size as compared to earlier. As the vascular system is a large and essential system of the body, it is largely affected due to any change in body; either by nature or by disease. The diabetes, hypertension, atherosclerosis, metabolic disorders [25] and tumors [26] are some diseases which affect the vascular systems and made it thick or thin. There have been several other efforts to investigate the utility of hand vein patterns for effective user authentication. In the following section, we present a brief review on the related prior work which is followed by the salient features of the proposed approach investigated in this paper.
Proposed SystemIn this paper, we develop a new hand vein authentication approach utilizing the structural similarity of hand vein triangulation and knuckle shape features. The block diagram of the proposed approach is shown in figure 2. The approach has been adapted to utilize palm dorsal images acquired from the, low-cost, contactless, near IR imaging. The main contributions from this paper [33] can be summarized as follows:(i) In this paper, we investigate the extraction and matching of hand vein structure using the key point triangulation. It many be noted that the low quality (visibility and clarity) of vein images does not guarantee same number of key points (from the same user) and therefore a weighted combination of four different types of triangulation is developed. Further details of this strategy can be seen section VII.(ii) The proposed method also investigates the utility of knuckle shape features since these features can be simultaneously extracted from the acquired images. The matching The image contours extracted from the acquired images are used for the image normalization and segmentation of region of interest (ROI) which is detailed in section II-IV. The automated extraction of hand vein map from ROI images is described in section V. The extraction and triangulation of local key points from the hand vain map is detailed in section VI and section VII respectively. The hierarchical matching scheme for the triplets is introduced in section VIII. The experiments and results from this work are presented in section IX which is followed by the discussion in section X and the main conclusions from this paper are summarized section XI.
II. Image AcquisitionThe acquisition of hand vein images using near IR imaging has been studied in [3], [11], [15]. In this work, a low-cost n...