A Novel and Fast Iris Location Algorithm Based on the Structure of Human Eyes
Weiqi Yuan, Zhonghua Lin, Xu Lu · 2006
In this paper we analyze eye images that we have based on the structure characteristics of eyes, we put forward a novel iris location algorithm. Firstly, get a point in the pupil by gray summing operator and find three inner boundary points that are not on the same line by the direction edge detection template designed by ourselves and locate the iris inner circle rapidly. Finally, find four outer boundary points that are not on the same line by the direction edge detection template designed by ourselves and locate the outer iris circle. 756 iris images comprising an iris image database were located, our algorithm only costs 1.59 minutes, average time of location reaches 0.126 sec at speed and 99.60% at accuracy. This algorithm is much faster than any other classical method such as Hough transform and the circle detection operator at speed. What is more, it ensures the robustness and the accuracy of location