Finger multibiometric cryptosystem based on score-level fusion

摘要

As a very promising approach, multibiometrics has been extensively studied to overcome the limitations of unibiometrics for high recognition performance in recent years. However, multibiometric systems always store multiple biometric templates for each user, which causes significant risks to user privacy and system security. In this paper, we address two critical issues in the design of finger multibiometric system, i.e., fusion methodology and template security. First, the score-level fusion framework using Fuzzy Commitment Scheme is proposed to protect multiple finger biometric templates of fingerprint, finger vein, finger knuckle print and finger shape traits. Second, the security of the proposed multibiometric cryptosystem is analysed on the basis of the correcting capacity of Error Correcting Code. The experimental results and security evaluation are conducted on a merged finger multimodal biometrics database. The comparative experiments show that the proposed finger multibiometric cryptosystem outperforms the unibiometric counterparts in terms of verification performance and template security.