A Robust Segmentation Method for Iris Recognition

In this paper, a new iris Segmentation method is presented. An iris recognition system acquires a human eye image, segments the iris region from the rest of the image, normalizes this segmented image and encodes features to get a compact iris template. Performance of all subsequent stages in an iris recognition system is highly dependent on correct detection of pupil-iris and iris-sclera boundaries in the eye images. In this paper, we present one such system which finds pupil boundary using image gray levels but uses Canny edge detection and Hough transform to locate iris boundary. Experiments are done on CASIA database of 756 iris images of 108 different persons with both left and right eyes images available per person. Experimental evaluation shows that the proposed system is accurate and efficient enough for real life applications. We are able to detect iris boundary almost 99.20% accurately with the proposed approach.