瞳孔亚像素边缘检测与中心定位

Subpixel edge detection and center localization of the pupil

为提高视线跟踪系统的分辨率和精确度,提出了一种亚像素精度的边缘检测与中心定位算法。利用设计的一维边缘检测算子对瞳孔像素级边缘点进行提取,然后利用边缘点小邻域内梯度方向上图像信息,进行亚像素边缘位置检测,最后经椭圆拟合确定瞳孔中心的精确位置及瞳孔形状参数。仿真实验验证了算法的有效性,提出的算法已被成功应用于实时视线跟踪系统中。

To improve the resolution and accuracy of a gaze tracking system,a sub-pixel edge detection algorithm is proposed in order to locate the center of the pupil accurately in IR image of the eye.With the proposed edge operator,the edge points of the pupil can be obtained by a 1D line search.Furthermore,utilizing the local information of the image in the gradient direction of the edge points,the algorithm can locate the edge points of the pupil in subpixel accuracy.And the algorithm can accurately determine the center and shape parameters of a human＇s pupil with an ellipse-fitting method.The experimental results indicate that the proposed algorithm achieved sub-pixel accuracy,which will significantly enhance the performance of the gaze tracking system. We have applied the algorithm to a real-time gaze tracking system.