Improving Iris Segmentation Performance Via Borders Recognition

This paper is focus on the analysis of the ability of iris segmentation method to process images with heterogeneous characteristics, simulating the dynamics of a non-cooperative environment. Aimed to improve the performance of iris segmentation method in non-cooperative environment, the authors introduce a new method. According to the different characteristics of the inner border and outer border of the iris, this new method segments iris by two steps: one step for inner border based on simple intensity operator, and the other step for outer border based on border recognition with support vector machine classifier. The performance of the proposed method and three selected traditional methods had been tested on the UBIRIS database. Compared with the traditional methods, the proposed method has more exactly accurate and less computation time on the image with poor quality. Also, the proposed method is higher robustness and less dependent on iris image conditions.