Vein pattern extraction based on vectorgrams of maximal intra-neighbor difference

摘要

In this paper, a vein pattern extraction method is proposed for biometric purposes. First, we utilize a maximal intra-neighbor difference (MIND) vector of all pixels in the original image to represent the relationship between each pixel and its neighborhood. Based on the MIND vectorgram (MINDVG), we define a maximal intra-neighbor vector difference (MIVND) as an index to unveil the preliminary vein pattern. Finally, we use an adaptive threshold to extract the venation pattern. The advantage of this method is that, by combining the features of vein imaging and the spatial properties of the MINDVC, the algorithm can efficiently overcome the negative factors of inhomogeneous thickness and blurry boundaries in vein imaging without preprocessing. Experiments on several images show that this method can directly extract intact and clear vein patterns with minimal noise. Therefore, the proposed algorithm has been validated in vein pattern extraction.