本文针对原有基于结构特征的虹膜定位方法提出一种有效的改进方法。本方法主要针对定位容易出现错误的外边界,利用虹膜的结构特征和数学手段,对得到的虹膜外边界点进行验证,必要时通过计算得到准确的外边界点。这样本方法解决了原有方法在眼睑和眼毛对虹膜的遮挡比较严重的情况下难以准确定位的问题。对已有的1 200幅虹膜图像(图像大小为640×480)组成的虹膜数据库进行实验,本文算法定位准确率达到99.08%,平均定位时间仅为0.8s,实验验证了该方法的有效性。

On the basis of the original iris location method based on the structure feature, an improved method is proposed. To the iris’ outer boundary, we use the feature of the iris’ structure and mathematical means to verify whether the iris outer boundary points are correct, and calculate the correct ones if necessary. Because of the disturbances of the eyelid and the eyelashes, the original iris location method based on the structure couldn’t locate the iris accurately, but the proposed method could effectively solve the problem. 1200 iris images (640×480 in size) comprising an iris image database are located in the experiment. The location accuracy of proposed algorithm reaches 99.08%, average time of location is only 0.8 s. The availability of proposed method is proved in the experiment.