基于灰度曲线几何特征的虹膜内边缘定位方法

Location method of iris inner boundary based on geometrical features of the curves of gray levels

为了解决以往虹膜内边缘(瞳孔边缘)定位方法算法复杂、计算量大和抗干扰性差的问题,提出了一种新的内边缘定位方法。研究了虹膜图像行、列灰度曲线的几何特征与瞳孔区域边界之间的关系,即:虹膜图像行、列灰度曲线上与瞳孔区域对应的部分有类似于“深井”的几何特征,利用它们之间的关系,通过寻找行列灰度曲线上“深井”的最大“井宽”定位瞳孔位置。实验结果表明此方法算法简单、计算量小、抗干扰性强,并且能够准确的定位虹膜的内边缘。

To reduce the complexity and calculation burden associated with locating an iris inner boundary, as well as to enhance noise rejection, a new method is proposed for locating an iris inner boundary based on the geometrical features of the curves of gray levels. First, the relationships between the geometrical features of the curves of gray levels for different columns and rows within the edges of a pupil were determined, with the gray levels having a geometrical character like a deep well. Second, the iris inner boundary was located based on the pointed relationships by finding the maximal width of the deep well. Experiment results show that the proposed method is simple, requires minor calculations, exhibits a strong ability to reject noise, and is accurate in locating the iris inner boundary.