Abstract: A novel subpixel location algorithm for elliptic center of circular targets image based on Zernike moment and curvature preserving is proposed.Firstly，pixel-level edges are detected and noises are filtered by multi-structure element multi-scale edge detection algorithm based on mathematical morphology.Secondly，a Zernike moment model for ellipse is developed and linked with curvature preserving to detect the subpixel edges.Finally，the subpixel elliptic center can be accurately obtained by fitting the edge points using least-squares fitting method.The experimental results indicate that the algorithm for circular targets center based on Zernike moment and curvature preserving has advantages such as faster computing speed and higher location accuracy，and it is very effective in high accuracy vision measurement.

摘要： 提出了一种基于Zernike矩和曲率不变的圆形标记椭圆图像中心的亚像素精确定位方法。首先采用多结构元多尺度形态学边缘检测算子提取椭圆图像的像素级边缘并滤除噪声，其次构造出椭圆图像的Zernike矩求解模型并结合曲率不变性计算出椭圆的亚像素边缘，最后利用最小二乘拟合对椭圆中心进行精确定位。实验结果表明：该方法具有计算速度快、定位精度高的优点，可用于高精度视觉测量。