Multi-unit Biometric Fusion in Fingerprint Verification

Abstract

The fingerprint verification system using many fingers is more efficient and reliable than using a single finger. Thus, the multiple use of several fingers and multiple impressions of the same finger can lead to an increase in accuracy. However, it will cause further inconvenience to the user and increase the overall verification time. Therefore, we combine only two fingers (multi-unit fusion) to improve the performance of the fingerprint verification system considering the processing time. We show that the performance can be improved by selecting a better quality fingerprint image of two fingerprints at a preprocessing stage. Also, we propose a new quality checking algorithm based on the directional field. Our experimental results show that when the better fingerprint selection is performed using the quality checking algorithm, there is a significant improvement in the performance of the fingerprint verification system.