An Efficient and Accurate Pupil Detection Method for Iris Biometric Processing

Abstract Iris-based biometric system is gaining its importance in several applications. Recent applications demand iris images to be processed as fast as possible with fewer number of computations to authenticate a person. Further, these applications are to be dealt with a large number of customers in security-critical domain, where it is important to ensure accuracy so that false confidences are avoided. As the source images are not guaranteed to be of good quality, it becomes a challenge to authenticate withalmost 100% accuracy in real-life applications. In this paper, we address these two problems and propose a technique to detect pupil boundary efficiently and accurately. We propose scaling and intensity transformation followed by thresholding and finding pupil boundary points. Scaling reduces the search space significantly and intensity transform is helpful for image thresholding. Experiments on Bath University, Multimedia University (MMU), UBIRIS and ICE 2005 iris image databases reveal that with the proposed approach, we able to detect pupil with accuracy rate almost 100% for Bath and MMU databases, 96% for UBIRIS database and 98% for ICE 2005 database.