Design and Analysis of Multimodal Biometric Authentication System using Machine Learning

Abstract

In this paper, a multimodal biometric identification system based on feature level fusion and machine learning techniques is described. The significance of this study relates to the combination of face and palm print for an individual identification. Machine Learning utilised to improve the performance of a multimodal biometric identification system. The performance evaluation is evaluated based on precision, recognition rate, equal error rate, and numerous evaluation metrics. The suggested multimodal system has an accuracy of 89.96 %, a false acceptance rate (FAR) of 3.32 %, and a false recognition rate (FRR) of 2.92 %. In order to arrive at this result, the multimodal system relies on score level fusion. It is demonstrated that a multimodal system may achieve high accuracy while using minimal FAR and FRR.