A modified Reed Solomon Error Correction Codes for multimodal biometrics recognition

Many existing real-world biometric applications are primarily unimodal. However, unimodal biometrics accuracy is often limited due to aging, and variations of interaction between user and sensor. In this paper, we propose a new multimodal authentication framework. The framework comprises of two main components which were Error Correction Codes and Weighted Score Level Fusion. Error Correction Codes is used to detect and correct the errors of two correlated features. On the other hand, Weighted Score Level Fusion is used to fuse the score of iris recognition and the face recognition to increases the level of accuracy performance of the biometric authentication. The dataset use for the experiment is self-established dataset (UTMIFM), WVU-IBIDC, UBIRIS version 2.0 and ORL face databases. The proposed framework achieves high accuracy, and had a high decidability index which significantly separate the distance between intra and inter distance.