Current Issue Cover
基于几何特征的虹膜定位算法

王成儒1, 胡正平1(燕山大学信息学院通信电子系,秦皇岛 066004)

摘 要
提出了利用虹膜边缘图象的几何特征以及圆相交弦的性质进行虹膜定位算法.首先根据虹膜图象的整体灰度分布信息,利用边缘检测算子提取虹膜的内边缘图象,然后利用圆的相交弦性质以及投票策略提取虹膜的中心,最后利用 Hough变换的半径直方图投票策略提取虹膜内外径.通过仿真实验的对比结果表明,该算法提高了虹膜定位的速度,同时对噪声具有一定的鲁棒性
关键词
Iris Location Algorithm Based on Geometric Features

()

Abstract
Iris location velocity and accurate is crucial in iris identification system. A iris location algorithm using geometric symmetry feature and two intersecting chord properties is presented. Based on this idea, edge detector and gray distributing information are used to extract iris interior edges , intersecting chord properties and voting scheme is used to search iris circle center. Then iris interior and exterior radius detected by using 1D radius histogram of Hough transform. At last the experiment results of iris location algorithms is showed, and the location time and accurate are compared with other method. Simulation results show that the method can improve location speed and isn't sensitive to noise effect.
Keywords

订阅号|日报