Development of a Two-Level Segmentation System for Iris Recognition Using Circular and Linear Hough Transform Algorithm

Recently, the use of biometric systems is of increasing significance in both public and private sectors. It has gradually replaced traditional system security. Iris recognition biometric systems have proved to be efficient at personal recognition with highest recognition accuracy. In this work a hybrid segmentation technique for iris recognition using circular and linear hough transform algorithms was developed. The hybrid segmentation technique was evaluated in terms of segmentation and recognition time. False Acceptance Rate (FAR) and False Rejection Rate (FRR) were the parameters used for comparison in the iris recognition. The system was implemented in MATLAB. After evaluating the performance of the hybrid segmentation techniques it revealed that the system was faster during the two stages. The averages segmentation time was 0.21089 seconds while average recognition time was 0.04113 seconds. The result shows the approach is efficient for the purpose of authentication.