Recent Advancements in Finger Vein Biometrics: A Review

For biometric purposes, many parts of the body are used for recognition, and specifically for vein recognition the palm, finger, hand dorsal, and wrist are used. There exist a couple of other biometric recognitions from various body parts such as the fingerprints, retina, face, footprints, ear, palm prints, retina, vein, etc., but among them, vein recognition is much more reliable and accurate. Specifically, finger vein recognition techniques have gained importance in the last decade with various methods and techniques using classical and deep learning approaches. However, there is limited use of finger vein recognition techniques using deep learning and especially for authentication purposes. For finger vein recognition, various feature extraction techniques and classification algorithms have been used. In this paper, different techniques used for feature extraction and classification from both traditional and deep learning perspectives of finger vein recognition along with the accuracy achieved and other performance metrics have been shown.