A new effective algorithm for iris location
Lijun Zhou,Yide Ma,Jing Lian,Zhaobin Wang
DOI: https://doi.org/10.1109/ROBIO.2013.6739727
2013-01-01
Abstract:Iris location is an essential step and an important part in an iris recognition system. However, traditional iris location methods often involve a large space of search, which is calculation wasting and sensitive to noise. And these methods adopt circular orientation to locate the pupillary boundary; it may lead to inaccurate location result and influence the subsequent feature extraction and recognition. To address these problems, this paper presents a precise iris location algorithm based on Vector Field Convolution (VFC, an improved Snake model) to improve the accuracy of iris location. Firstly, obtaining the iris area completely include the inside and outside boundary from an original iris image, then using minimum average grey value method to determine initialization contour of VFC model automatically, so as to locate an iris inner boundary precisely under the internal and external force of active contour. At last, we adopt the improved Daugman algorithm to locate the iris outer boundary that relatively contains little texture information. Experimental results show that the location accuracy of this method is higher, the iris inner edge location is much closer to the real boundary, the result of location have been improved significantly.