A Novel Finger Vein Pattern Extraction Approach for Near-Infrared Image

Abstract

Biometric authentication has been extensively studied for many years and attracted much attention due to its large potential security application. Finger vein is more stable and more difficult to fake than on-touch feature such as fingerprint, palm print and face. In this paper, we proposed a novel finger vein pattern extraction method based on low quality vein image, which is generated by common CCD camera absorbing near-infrared light. Firstly, local threshold method is adopted to give initial points set by probability; Secondly, a modified line tracking algorithm is performed, and a new probability map is obtained; Afterward, we synthesize a thorough probability map and create the corresponding gray map reflecting the probability of vein point. Finally, a directional neighborhood analysis is made on the previous probability map to extract vein pattern. Experimental results illustrated that the proposed method can achieve better results than existing methods.