Simultaneous acquisition of near infrared image of hand vein and pulse for liveness dorsal hand vein identification

Abstract

Abstract In this paper, a living dorsal hand vein identification method based on acquisition near infrared image of dorsal hand vein and pulse at the same time and recognition algorithm is proposed. An experimental prototype device is designed and built with 850nm LEDs and a CCD camera, which realizes the simultaneous acquisition of dorsal hand vein and finger pulse information. Then, the characteristics and differences of near-infrared signals between prosthesis and living body over time were studied with 20 copies of dorsal hand vein images and 20 living bodies, which verified the effectiveness of the rapid detection method of photoelectric volume living pulse. On this basis, 900 vein images of 45 living bodies are captured, and the influence of palm position and the algorithm to improve the accuracy of vein recognition are studied. Combing with rotation and translation transformation, PLBP and SVM algorithm, the recognition accuracy is 99.33%.