Improvement of Iris Images Location and Segmentation Algorithm
Wang Yan-nia · Jiguang zazhi · 2015
An improved algorithm is proposed in this paper based on iris image location and segmentation to improve the recognition rate of iris identification system. Firstly,binarization method is used to segment the pupil and then the original images been shrinked. Secondly,the edge is detected via an improved Canny operator,after that a least squares fitting is used to curve fitting. Finally,both up and down sides of the eyelids is eliminated by Radon transform and the feature point of iris region is obtained. Experiments based on CASIA1.0 and CASIA4.0 iris images on the platform of Matlab. Experiment results show that this method not only could significantly reduce the location time,but also achieve 98. 82% of location precise rate,which with practical application value.