A Novel Binary Method for Fingerprint Optical Coherence Tomography Images

Abstract

This paper described a novel binary method for inhomogeneously lighted fingerprint optical coherence tomography (OCT) images. To reduce the light effect problem, a block size is calculated by the median value of the length values of white pixels between black pixels in the initial binary result and used to rescale the gray values of the input image. The rescaled gray image is used to calculate the binary candidate image and the binary gradient direction image. The final ridges are calculated by using the relationship between the binary candidate image and the binary gradient direction image.