作　　者： ; ; ;

机构地区： 华中科技大学机械科学与工程学院仪器科学与技术系

出　　处： 《计算机工程》 2002年第12期129-130,共2页

摘　　要： 如何有效地确定虹膜在图像中的位置，是虹膜识别中需要解决的关键问题之一。在以前的粗定位与精定位相结合的两步虹膜定位方法的研究基础上，提出了一种改进的虹膜定位方法，它集中体现在虹膜粗定位方法的改进上。改进后的虹膜粗定位的基本思想为：首先在虹膜图像中选择互不相邻的若干行，求得每行的灰度差分曲线；然后根据灰度差分曲线上相邻几个点的累加和的大小，寻找虹膜或瞳孔的边界点；最后，由这些边界点拟合出虹膜及瞳孔的边界。实验结果表明，改进的虹膜定位方法更为快速、稳健。 How to find the true iris position in an iris image is a key problem to be solved in iris recognition. An improved method of iris location is presented based on the previous two-step iris location combining coarse location with fine location and the improvement focuses on coarse location. The principle of the improved coarse location approach is as followings: Firstly, the grey scale difference curve for each row can be gotten after several rows are chosen which aren't adjacent each other; And then some boundary iris or pupil points are found according to the value of the cumulative sum of several adjacent points on the grey scale difference curve; At last, the iris or pupil boundary is fitted from the boundary points. Experiments show the improved method is faster and more robust.

关 键 词： 虹膜定位 灰度差分曲线 虹膜图像 虹膜识别 边界拟合

领　　域： [自动化与计算机技术] [自动化与计算机技术]