虹膜图像内外边缘定位算法研究

Study on Algorithm for Iris Image Inner and Outer Boundaries Location

虹膜定位是虹膜识别过程中的重要环节,定位的速度和精度决定了整个虹膜识别系统的性能,精确的虹膜定位是有效进行虹膜识别的前提,所以一种简洁有效的虹膜定位算法是至关重要的。针对现有的虹膜定位算法的局限性,提出了一种新的虹膜定位算法。首先从虹膜图像的灰度直方图入手,分析虹膜图像的结构及其灰度分布,先用二值化的方法分离出瞳孔,再用数学形态学方法对其进行膨胀和腐蚀运算,然后确定瞳孔的圆心和半径,对内边缘进行精确定位。利用虹膜定位的先验知识,在缩小搜索范围的基础上,再用形态学算法与新的Hough变换相结合来进行虹膜图像外边缘的定位,确定外圆圆心及半径。实验结果证明了该算法的有效性和可行性。

Iris location is an important part of iris recognition process,and its speed and accuracy determines the performance of iris recognition system.The precise iris location is the premise of iris recognition,and thus an effective iris terse location algorithm is of the utmost importance.Based on study and analysis of the existing iris location algorithm and aiming at the limitation of the existing iris algorithm,a new method to localize the iris is proposed.First,the iris image structure and gradation distribution is analyzed in terms of iris image gradation histogram,and the pupil is separated with the binaryzation method.The inflation and erosion of mathematics morphology method is used to locate the inside of the edge by defining the center and radius of the pupil.The apriori knowledge of iris location is used to reduce the search range and locate outside edge of the iris image with the morphology algorithm in combination of the new hough transform.The method of edge detection based on mathematical morphology and new Hough transform is used to locate the outer edge of the iris,thus the center and raduis of outer annulus is obtained.The experimental results show that this algorithm is effective and feasible.