摘要:研究人眼虹膜识别问题,因实际虹膜内边界并不是标准圆,引起识别精度差,影响有效的特征提取.传统利用圆模板定位的算法存在瞳孔遗留或纹理损失且定位时间长等问题,为提高虹膜定位精度,降低识别时间,提出了一种新的虹膜识别算法.首先对图像进行去除光斑等预处理,将含有虹膜图像的圆环变换为极坐标系下的矩形,在矩形坐标上以点、线检测确定虹膜轮廓,并对EMD提取纹理分布特征,根据比对距离寻找每个待测样本的K个近邻,以简单投票决策输出识别结果.基于CASIA虹膜图像库进行仿真,结果表明,识别率高达99%,并明显降低了识别时间,使虹膜定位可有效提升识别精度.%In iris recognition, as a large amount of experiments show, the inner edge of iris is not an ideal circle, thus edges may incorporate some information that does not belong to the iris or some iris texture information is cut though traditional localization method using round template. To improve the accuracy of iris location, reduce the recognition time, this paper developet a new iris recognition algorithm. Firstly, the light pot within the pupil was filled in the original image, then the image was unfolded into a rectangle and the circle detection was substituted bythe point and line detection in the rectangle image to find the inner and outer edge. Secondly, texture features were extracted by EMD. Thirdly, the K nearest neighbors ( KNN) of each test sample were found based on distance of Ma-halanibis. Lastly, recognition results were decided by majority voting method. The recognition accuracy of simulation experiments based on CASIA iris image database is up to 99% and requires less running time. The results show that compared with circle template, rectangle convention locates the iris more accurately, and effectively raises the recognition accuracy.