光学测量中椭圆圆心定位算法研究

Algorithm of centre location of ellipse in optical measurement

提出了一种椭圆圆心定位算法。先对图像中的椭圆目标进行粗定位,获取椭圆的粗定位信息,再用Canny算子对图像中的感兴趣区域进行边缘检测。根据椭圆的粗定位信息以及图像的边缘灰度分布特征,用高斯曲线拟合的方法求得精确的亚像素边缘点的坐标。针对亚像素边缘点中出现的"孤立点"以及噪声点,分别用曲率滤波和均值滤波的方法加以滤波,将滤波后的亚像素边缘点用最小二乘法加以拟合,求得精确的椭圆圆心数据。实验证明了该算法的定位精度和抑制噪声能力,通过测试算法运行时间,证明该算法具有很好的实时性。

A new algorithm for obtaining the centre of ellipse is brought forward. Some information about an ellipse is obtained by the coarse-location, and the edge points are detected by using Canny operator in the region of interest. Then the accurate edge points are acquired by Gauss curve fitting based on the information of the ellipse and the gray-distribution characteristics of the edge points. Finally, the filtering methods based on the curvature fi!tering and mean-filtering are carried out respectively to filter the isolated points and the noisy points, and the accurate centre of ellipse is obtained by least square method. The result proves that the new algorithm locates the centers of ellipses accurately and meets the need of real-time measurement.