Informatics Publishing Limited

Abstract

Single biometric cannot provide much accuracy as compared to multibiometric, so multibiometric is preferred for real world personal identification application that needs high standard security. Multibiometric combines two or more single biometrics in order to get the higher accuracy. However, among various biometric technology palmprint identification has been proved more efficient and easy to implement. This paper reveals the idea to perform the multibiometric by combining the feature of the left and right palmprint.  This framework combines three kinds of scores generated by left and right palmprint images to perform matching score level fusion. Initially the score is generated from the left palmprint images followed by right palmprint images and finally the left palmprint is compared with the reverse right palmprint to obtain the third score. All the three scores are combined and based on final score the subject is authenticated.

The nature of the left and right palmprint images is carefully taken into account by proposed algorithm, which allows it to properly exploit the similarity between the left and right palmprint images. Moreover, as compared to the previous palmprint identification method, the proposed weighted fusion scheme allowed perfect identification performance.

Keywords

Full Text