基于灰度曲线法及改进Hough变换的虹膜定位算法

Iris location algorithm based on intensity curve and improved Hough transform

虹膜定位是虹膜识别系统的基础和关键,定位的速度和准确率直接影响着虹膜识别算法的性能.针对虹膜图像的灰度分布特征和近似圆形的几何特征,提出一种基于灰度曲线法和改进Hough变换的虹膜定位算法.首先,利用灰度投影法粗略定位瞳孔中心;其次,利用灰度曲线法精确定位瞳孔中心及虹膜内边缘;最后,利用改进的Hough变换检测虹膜中心及外边缘.仿真实验结果表明,该虹膜定位算法内存空间开销小,速度快.

Iris location preprocessing plays a key role in iris recognition. The speed and accuracy of location determine the performance of the iris recognition algorithm. According to the distribution of gray scale and geometrical characteristics of human eyes, an iris location algorithm was designed based on the intensity curve and an improved Hough transform. First, the intensity mapping was proposed to coarsely locate the center of pupil. Second, the intensity curve was used to precisely obtain the center of pupil and the inner boundry of the iris. Finally, an improved Hough transform was applied to precisely detect the outer boundry of the iris. The simulation results showed that the proposed algorithm consumes much less storage and computation cost than existing algorithms.