University of Puerto Rico
Human iris provides a unique structure suitable for non-invasive biometric assessment. In particular the irises are as distinct as fingerprints even for twins. In this paper a system for person identification is presented that uses a technique of localization, alignment, feature extraction, matching the features of irises and finally the decision regarding the degree of match based on hamming distance. A CASIA iris database of iris images has been used in the implementation of the iris recognition system. The results show that proposed method is quite effective.