Analysis of different techniques for score level fusion in multimodal biometrics

Abstract

Fusion in the multimodal biometric system is a very important part of the authentication process. Two modalities are combined to form a single authentication factor so selection of fusion method is also need very attention. As there are different level of fusion so in this work, score level fusion is used to analyze multimodal biometric system’s performance. This work uses two data sets CASIA and IITD and firstly texture features of both iris and fingerprint will be extracted. Then theses texture feature will further used to calculate score for both modalities and fuse using SUM, PRODUCR and MAX methods. Performance of all three methods has been analyzed in terms of FAR, FRR and accuracy.