Multibiometric Authentication System Using Slap Fingerprints, Palm Dorsal Vein, and Hand Geometry

摘要

In this paper, a multibiometric system that fuses slap fingerprints, palm dorsal vein, and hand geometry for accurate person authentication is proposed. The proposed system simultaneously acquires slap images and infrared (IR) hand dorsal image from which slap fingerprints, palm dorsal veins, and IR hand geometry are extracted. Simultaneous acquisition reduces the acquisition time and helps to improve user acceptability. The slap segmentation is accomplished using the knowledge of the approximate finger location and hand type (either left or right) obtained from the simultaneously acquired IR hand dorsal image. Multibiometric fusion based on serial methodology has been proposed for consolidating the matching scores of slap fingerprints, palm dorsal vein, and IR hand geometry. It ensures better performance in terms of accuracy and time. Experimental results indicate that authentication using slap fingerprints can be improved by incorporating the knowledge from IR hand images. Further, it depicts that the fusion of slap fingerprint, palm dorsal vein, and IR hand geometry can help to achieve the best performance.