Vascular Finger Vein Image Segmentation Using Bifold Contrast Limited Enhancement

Abstract

Image enhancement is never ending area of agile research. Standardizing the image quality is always a challenge in the imaging research field to achieve optimal imaging. One of the popular visual enhancement method widely used in medical image enhancement named as CLAHE (Contrast Limited Adaptive Histogram Equalization), in which contrast of the local regions are increased, there by enhance the local information of an image.

This paper proposes a scheme to extract the finger veins from the captured image background surface using bifold CLAHE technique. We apply CLAHE on finger ROI followed by finger vein ROI to enhance the finger vein pattern, then apply binarization based on Otsu thresholding in Canny edge detection to obtain optimum finger vein edges. Experimental results show that, for a given set of input images, our proposed bifold CLAHE where N and CL are calculated experimentally outperforms in terms of contrast enhancement there by containing the more image information compare to legacy methods.