一种有效的虹膜定位及睫毛检测方法

提出一种有效的虹膜定位及睫毛检测方法。通过把眼睛图像中分割成小的矩形区域，利用找到的这些矩形区域像素平均最小值把眼睛图像进行二值化，找到虹膜区域的内边界；以瞳孔的质心为参考点，在其左右的扇形区域内分别使用修改后的Daugman的检测算子，找到像素值变换大的位置，进而定位出虹膜的外界；使用Gobor滤波器和窗口移动法对睫毛进行有效的检测。通过对大量虹膜图像的实验表明，该方法取得了非常好的结果。

Efficient algorithm on iris image localization and eyelash detection

An efficient algorithm on iris image localization and eyelash detection is presented.Firsdy,an eye image is divided into many small rectangular regions.Image binarization is implemented taking the smallest average intensity of the regions as a threshold.The inner boundary of the iris region is found.Then the centroid of the pupil is taken as a reference point.In the left and right sector regions of the centroid,the modified Daugman's detecting operator is utilized to search the sharp variation position of the intensity.Then the outer boundary of the iris region is located.At last,eyelash is detected by using Gobor filter and the shift window.After doing a great deal of experiments,the results show that the proposed method has an encouraging performance.