A New Technique for Fast and Accurate Iris Localization
Kavita Khobragade, Babasaheb Ambedkar · 2014
Abstract- Iris Recognition is an emerging technology in today’s world; in both research and practical applications. The human iris contains rich texture information. This texture information is highly stable and distinct biometric characteristics. Iris biometry is used to recognize any individual in a natural and intuitive way. Iris localization is the most important step in Iris Recognition System and it determines the accuracy in feature extraction and matching. This paper proposes a new, fast and accurate algorithm for iris localization. In this proposed iris localization technique, three points were selected to detect the center of the pupil, inner and outer boundaries. The verification performance of the proposed algorithm is validated and compared with other algorithm by using CASIA version 1.0, CASIA version 3.0 Interval, CASIA version 4.0 Interval and MMU version 1 and 2 databases. Few iris images of PALAKY database were also tested with the proposed algorithm. The experimental result on the CASIA version 4.0 Interval demonstrates that the proposed method has better performance and it improves speed and accuracy for iris localization.