Vulnerability Assessment and Presentation Attack Detection Using a Set of Distinct Finger Vein Recognition Algorithms

Abstract

The act of presenting a forged biometric sample to a bio-metric capturing device is referred to as presentation at-tack. During the last decade this type of attack has been addressed for various biometric traits and is still a widely researched topic. This study follows the idea from a previously published work which employs the usage of twelve algorithms for finger vein recognition in order to perform an extensive vulnerability analysis on a presentation at-tack database. The present work adopts this idea and examines two already existing finger vein presentation attack databases with the goal to evaluate how hazardous these presentation attacks are from a wider perspective. Additionally, this study shows that by combining the matching scores from different algorithms, presentation attack detection can be achieved.