Output details
11 - Computer Science and Informatics
University of Northumbria at Newcastle
An effective and fast iris recognition system based on a combined multiscale feature extraction technique
<19>State of the art iris recognition systems are inefficient when iris images are distorted or noisy especially when taken under varying conditions. This paper proposes a new multiscale edge detection approach as a pre-processing step to efficiently localise the iris from distorted/noisy images. This has led to a compact and efficient feature vector resulting in a system which outperforms existing systems especially in the case of noisy and blurred images. This work lead to an invitation for research collaboration from Qatar University and Qatar Security Office in 2009 resulting in a funded project from Qatar Science Foundation (NPRP 09-984-1-128, 2009).