Multi-Modal Iris and Retina Recognition System using Discrete Hidden Markov Model based Score Fusion Technique

Abstract

The contribution of this work is to propose a model of multi-modal iris and retina recognition system where Hidden Markov Model based score fusion technique has been used for decision fusion approach. Two different uni-modal techniques such as iris recognition and retina recognition outputs are combined by using baseline reliability ratio based score fusion technique. CASIA iris dataset and DRIVE retinal dataset have been used to acquire the iris and retina images for the proposed system. Effective image pre-processing techniques have been used to process the iris and retina images. Reliability of each uni-modal i.e., iris or retina has been measured from the output of Hidden Markov Model classifier. The reliabilities of these two uni-modal systems are integrated to find out the final result. Experimental results’ performance analysis shows the versatility of the proposed iris and retina recognition based multi-modal system over each uni-modal system.

Keywords- Multi-modal iris-retina recognition system, score fusion technique, hidden Markov model, reliability measurement

Cite this Article

Md. Rabiul Islam, Md. Fayzur Rahman. Multimodal Iris and Retina Recognition System using Discrete Hidden Markov Model based Score Fusion Technique. Current Trends in Information Technology. 2015; 5(2): 7-12p.