Iris-Based Human Verification System: A Research Prototype

The biometric person authentication technique based on the human iris is well studied to be applied in any access control system requiring a high level of security. In this paper a system for personal verification based on iris patterns is presented. This includes a problem of biometric systems, required knowledge of the anatomy of the human eye, what is verification based on the iris patterns and algorithm operating principle. In the middle part, we present the verification algorithm, which consists of image processing to obtain iris information, iris normalization, feature extraction, and person verification. Gabor filter was used for feature extraction. From such result iris bit template sequence is encoded. Then the Hamming distance is calculated from the iris template, which gives the estimate of the match in the verification process. The results of the algorithm were obtained on CASIA iris database, where a decision threshold for the Hamming distance was set to 0.427, which gives 0% false acceptance rate and 11.584% false rejection rate on the test set. As we were developing a prototype system with the goal to build a foundation for possible improvements and for a real test system, we were very satisfied with the achieved results.