PERSONAL IDENTIFICATION USING MULTIBIOMETRICS SCORE-LEVEL FUSION

Amritha K.C, S.Velmurugan

Abstract: A combination of more than one biometric is called multimodal biometric system. Here proposing an identification system by combining the feature of face and signature at matching score level. Thus, it is expected that spoofing two or more independent data would present more difficulty as compared to just one biometric data. With the hypothesis that impossible to spoof attack the physical quality and behavioral characteristic simultaneously, face and signature selected as the modalities of the biometric to be merged. The fusion, at the score level provides a matching score indicating the proximity of the feature vector with the template vector. These stacks can be blended with a simple sum rule. Storage of database in the form of these scores gives protection from hacking the databases. The results show that the error rates for this bimodal identification system is less than a single face identification system.GAR and FRR are used to evaluate the system performance.

Keywords: Face, Signature, Genuine Acceptance Rate, False Rejection Rate

DOI: https://doi.org/10.15623/ijret.2014.0319145