Implementation Of A Multimodal Biometrics System Based On A Feasible Fusion Of Single-Matching Methods

Abstract

A biometric system in which evidence of identity is based only on a certain biometric trait. Multimodal biometric recognition is a new trend that is growing rapidly. Typical multimodal biometric authentication solutions use more memory, have poor response rates, and have greater adoption and operating costs. This article presents an innovative structure for multimodal biometric identification systems that is adaptable to any form of biometric, resulting in a reduced memory footprint and speedier implementation. The suggested framework is validated by the building of a fusion system based on a single main component matcher for faces, fingerprints, and palm prints. The system is extensively tested using fusion in identification mode. The results indicate that the system outperforms a single unimodal system in terms of precision and is also equivalent to a standard multimodal system to achieve an outcome.