An algorithm for finger-vein segmentation based on modified repeated line tracking

Abstract

Image segmentation is an important step for finger-vein identification technique. However, it is difficult to extract precise details of the image because of the irregular noise and shades around the finger-vein. The repeated line tracking algorithm achieves good segmentation performance for low quality images of finger-vein, but it has some drawbacks such as low robustness and efficiency. In this paper, a modified repeated line tracking algorithm is proposed for image segmentation of finger-vein. Firstly, we propose a segmentation method called threshold image to execute rough segmentation and obtain binary and skeleton image of fingervein. Secondly, the width of finger-vein is estimated based on the binary and skeleton image. The parameters are revised according to the width. Then, the modified repeated line tracking algorithm is executed to figure out the locus space of finger-vein based on the revised parameters. Finally, processing results are obtained by using Otsu algorithm which executes exact segmentation on the locus space. Experiments show that the proposed algorithm is more robust and efficient than traditional repeated line tracking algorithm.

Finger-vein recognition with modified binary tree model

“…1 Framework of the proposed method. There are four stages involved: 1 rough selection: execute simple feature matching between two MBT models, to reduce the number of samples in dataset and improve the efficiency of entire recognition process of 1:N recognition mode, 2 model correction: align the MBT models for matching, to reduce the influence by translation and rotation of finger, 3 segment matching: divide finger-vein branches into segments without overlap and calculate the matching errors of all segments independently, to differentiate the matching errors caused by segmentation error and local difference, and 4 comprehensive judgment: divide the matching errors into three parts and judge the matching error caused by segmentation error and local difference independently, to boost the performance of finger-vein recognition degraded by these factors Neural Comput & Applic median filter [24]; image segmentation adopts MRLT method [25]; image smoothing makes the use of open morphological operators [24]; hit-or-miss morphological operators [24] are used in image thinning; burrs are removed according to burrs' length. The details are described in Ref.…”

Finger-vein pattern restoration with Direction-Variance-Boundary Constraint Search

“…However, it is difficult to extract the precise details of the depicted finger-vein patterns because of the low quality of finger-vein images . Although existed algorithms (Zhang et al, 2006;Yu et al, 2008;Ding et al, 2005;Mulyono and Jinn, 2008;Vlachos and Dermatas, 2008;Miura et al, 2004;Liu et al, 2013) can extract most finger-vein patterns, some branches of these finger-vein patterns always break. Therefore, finger-vein patterns extracted from different images of the same finger are far different from each other, and hardly match with each other in the process of features matching.…”

“…However, there are a few restoration methods at feature level. The existed methods only use some filtering technology such as mathematical morphology and median filtering, to restore the segmented images (Yu et al, 2008;Liu et al, 2013;Yang and Shi, 2014). These methods restore veins with a distance metric.…”

Finger Vein Recognition Using 2DGabor Combined With OECA

“…The contour point feature is the earliest applied to finger vein recognition to extract the bifurcations point and the endpoints as the classification information [5], and then the SIFT (Scale-invariant feature transform) feature points are introduced to enrich the robustness against rotation [6], but the contour point feature will lose a lot of information. Finger vein line can reflect the vein topology, linear tracking method [7], regional growth method [8], curvature method [9] is often used to extract the vein line. LBP (Local Binary Pattern) is widely used in finger vein recognition as a class of methods with strong texture description ability [10], but at the same time it is insensitive to the changes in light intensity, Gabor filter banks can extract the global and local information of the veins, and the obtained texture features have good texture description ability and anti-noise ability [11].…”