提出了基于单位扇环灰度的虹膜定位算法。利用投影法分割出一个包含瞳孔的矩形区域, 通过最大类间方差法确定该矩形区域的阈值, 完成瞳孔的分割。对于虹膜外边界的定位, 需要基于瞳孔中心分割出某一方向上的扇形区域, 以5个像素作为扇形区域的步长, 计算每个扇环的平均灰度值, 根据灰度变化情况就可以确定该方向上的虹膜外边界点。其他方向上的边界点也通过此方法确定。对这些边界点进行筛选并进行圆的拟合, 最终实现虹膜的定位。实验结果表明, 采用该方法所分割的虹膜图像可以获得良好的效果, 具有良好的应用和参考价值。

A method of gray value of unit sector area is proposed to solve the problem of iris localization. A rectangular area containing pupil is segmented by gray-scale projection algorithm, the OTSU algorithm is used to determine the threshold of this rectangular area and segmented the pupil. With respect to outer location in a certain direction, a sector area in the direction needs to be segmented according to the pupil center. The gray value of each concentric sector ring is calculated by setting steps to 5 pixels,and the abscissa values of the position where the maximum change in gray value is calculated as the radius in this direction.The location of boundary points in other directions are also applied by this method. These points are filtered and used to fit outer boundary of iris. The results show that iris images segmented by this method can achieve a good result with good application and reference value.