Novel iris recognition method
Qi Wang · Guangdian gongcheng · 2005
A novel iris recognition method based on multi-texture feature fusion is presented in the paper. The extraction of the texture features with different resolutions in different directions are carried out by performing Gabor wavelet transform for iris image and the extracted features are used as global features of iris. The local features of iris are extracted by applying gray scale level Co-Occurrence Matrix (COM) to the filtered sub-window image. The global features and local features will be recognized and sorted by weighting Euclidean distance and the minimum distance, respectively. A sorting method with Fuzzy Inference System (FIS) feature fusion is designed to improve the robustness of iris recognition. Experimental results show that the method is effective and feasible with recognition ratio up to 98.5%, False Acceptance Rates (FAR) lower than 0.1% while False Rejection Rates (FRR) keeping 1.4%.