Face and Iris based Multimodal Person Identification using Likelihood Ratio Score Fusion Approach

Abstract

The aim of this paper is to propose a multimodal biometric system where likelihood ratio score fusion technique has been used to combine face and iris uni-modal biometric system outputs. Active Shape Model (ASM) has been used to extract the facial features and standard methods have been applied to effectively extract the human iris features. To evaluate the output of each individual modality of face and iris, Discrete Hidden Markov Model (DHMM) based classification technique has been applied. Over various standard decision fusion approaches, likelihood ratio score fusion technique has been used to measure the performance of the proposed system. To evaluate the performance of the proposed system, ORL face database and CASIA-IrisV4 database have been used together for face and iris identification performance, respectively. Experimental results and performance analysis show the superiority of the proposed multimodal system with the comparison of individual face and iris identification performance.

Cite this Article

Md. Rabiul Islam, Md. Abdus Sobhan, Rizoan Toufiq. Face and Iris based Multimodal Person Identification using Likelihood Ratio Score Fusion Approach. Journal of Computer Technology & Applications. 2015; 6(3): 24–31p.