Iris recognition using radon transform and GLCM

Iris recognition for some time now has been a challenging exercise. This perhaps is due to the use of inappropriate descriptors during the feature extraction stage. In this paper, a Radon Transform is used as an iris signature descriptor. Blood vessels are segmented from iris image. After blood vessel segmentation, the radon transform is applied on the segmented image. The GLCM, Gabor and Local Binary Patterns are used for feature extraction. Well known SVM classifier is used for classification of the iris data. The performance of the system is evaluated on DRIVE and High Resolution Image Databases. The system performs better for the radon transform signature. This proposed approach of iris recognition using blood vessel segmentation is robust and secure and has the ability to recognize retinal images from the photographs of the known iris images. The system is more efficient in terms of accuracy as well as time complexity. The GLCM features when applied on Radon signatures gives improved results for both datasets of DRIVE and HRF.