Abstract

In this paper we introduce an algorithm to analyze the human iris, long-range iris recognition software has been developed  to be more user-friendly and create an economic way to the identification. Our algorithm centralized on pupil detection, and by using estimated ranges we omit the other regions to create more efficient search space. The final decision on iris region detection provided by Hough Transform. We use the Gaussian method to create a refined mask which has an important role in the matching process. To extract efficient features of iris regions and matching we used SIFT algorithm, Results on CASIAV4-at Distance shows %93 as verification Rate.