Fingerprint Segmentation Using Scale Vector Algorithm

Abstract

Pre-processing part in fingerprint image recognition system is fingerprint image segmentation. Even after the years of research in the field of image segmentation, it is still a challenging task. An image segmentation algorithm is proposed, named as Adaptive (scale) and Orientation (vector) algorithm, for effective segmentation of fingerprint images. The basic idea of the proposed algorithm is originated from the total variation models, along with two features of fingerprints, namely, scale and vector. It decomposes a fingerprint image into two layers: noisy and texture. Proposed segmentation algorithm is experimented and analyzed for two different fingerprint images. PSNR is the measuring parameter for checking the efficiency of algorithm. This algorithm is compared with adaptive and directional algorithm individually and concludes its better efficiency.