ABSTRACT
Professor J. Daugman sparked the idea of iris recognition in the biometrics world and it has been commercially used in various industry. Iris image is transformed into series of binary format called as IrisCodes and saved in the database. The stored IrisCodes are compared with a real IrisCodes at the matching process to find similar binary bits. Iris recognition involves image acquisition, segmentation, normalization, extraction and comparison to generate the IrisCodes. Furthermore, circular segmentation techniques were used such as Integral Daugman Operator, Hough Transform, Principal Component Analysis and wavelets. However, the IrisCodes produce a flipping bits situation and unintentionally accept the impostor as the real user to access the system. Another situation is the genuine user who gradually known as the impostor user due to high noise rate in the iris template. The biggest challenge of the first phase of iris recognition is to find the similar bits in IrisCodes during matching process due to problems of occlusion, distortion, pupil size and aging.
