Finger-Vein Biometric Recognition: A Review

Abstract

Finger-vein-based biometric technology has received wide attention and has made some positive achievements in improving the performance of personal identification. This paper intends to provide an overview on the existing finger-vein-based biometric methods from the four steps of finger-vein recognition methods, including image acquisition, image preprocessing, feature extraction, and matching. After brief introduction of image acquisition and preprocessing, the applications of feature extraction method are summarized as: template-based method, representation-based method, and learning-based method. Moreover, deep learning-based methods have been analyzed and discussed. To review and summarize these methods, case studies of the finger vein database are introduced to verify their effectiveness. Meanwhile, this paper discussed the challenges and prospects of finger vein recognition research, especially the three-dimensional finger vein recognition research, to guide the future research direction in this field.