基于并查集和边缘检测模板的非理想虹膜定位

非理想虹膜图像往往存在虹膜边缘模糊、灰度变化不均匀、位置偏移及光斑干扰等问题, 这些问题的存在会在一定程度上影响虹膜内、外边界定位的准确率. 针对这个问题, 本文提出采用并查集和边缘检测模板的方法来对非理想虹膜进行内、外边界定位. 内边界定位是首先采用并查集方法完成瞳孔区域粗定位, 然后采用Hough变换对瞳孔进行精确定位; 外边界定位是先利用一系列边缘检测模板大致确定外圆的位置, 然后再依据外边界附近圆环内边缘点的密度来最终完成外边界的精确定位. 实验结果表明, 对于非理想虹膜图像, 本文方法的定位正确率和定位速度均高于其他同类方法.

A non-ideal iris localization method based on Union-Find-Set and edge detection template

Non-ideal iris image tends to have some noise interferences, such as edge blur, uneven gray level, position shifts , specular-reflections and so on, which will affect the localization accuracy rate of the inner and outer boundary of the iris.To solve this problem, a novel iris location approach based on Union-Find-Set and edge detection template was proposed in this paper. Firstly, Union-Find-Set was implemented to coarsely detect the pupil area, and then Hough transform was used to localize the inner boundary precisely. Secondly, a series of edge detection templates were employed to coarsely locate the outer boundary, after that, accurate outer boundary localization was carried out based on the density of edge points in a circular ring near the outer boundary. The experimental results show that the proposed method outperforms the state-of-the-art method not only on localization accuracy rate but also on localization speed.