opencv中的椭圆拟合

首先贴一个最简单的程序：访问：https://blog.csdn.net/guduruyu/article/details/70069426

//创建一个用于绘制图像的空白图    
cv::Mat image = cv::Mat::ones(480, 640, CV_8UC3);  
//设置蓝色背景  
image.setTo(cv::Scalar(100, 0, 0));  //输入拟合点    
std::vector<cv::Point> points;  points.push_back(cv::Point(200, 240));  
points.push_back(cv::Point(300, 400));  
points.push_back(cv::Point(400, 360));  
points.push_back(cv::Point(500, 300));  
points.push_back(cv::Point(500, 200));  
points.push_back(cv::Point(300, 150));  //将拟合点绘制到空白图上    
for (int i = 0; i < points.size(); i++)  
{  cv::circle(image, points[i], 5, cv::Scalar(0, 0, 255), 2, 8, 0);  
}  //获取拟合椭圆的外包围矩形  
cv::RotatedRect rotate_rect = cv::fitEllipse(points);  
//绘制拟合椭圆  
cv::ellipse(image, rotate_rect, cv::Scalar(0, 255, 255), 2, 8);  cv::imshow("image", image);  
cv::waitKey(0);

其次，这里有关于opencv椭圆拟合的各种定义：https://blog.csdn.net/suky520/article/details/18601307

opencv中的椭圆通过RotatedRect来定义，其center(块中心(x,y)), size(宽和高,宽小于高), angle（旋转角）分别对应椭圆的中心，短轴和长轴，旋转角度。

         RotatedRect box = fitEllipse(pointsf);  //把那些长轴与短轴之比很多的那些椭圆剔除。  if( MAX(box.size.width, box.size.height) > MIN(box.size.width, box.size.height)*8 )  continue;  //绘制轮廓  drawContours(cimage, contours, (int)i, Scalar::all(255), 1, 8);  //绘制椭圆  ellipse(cimage, box, Scalar(0,0,255), 1, CV_AA);  //绘制椭圆  // ellipse(cimage, box.center, box.size*0.5f, box.angle, 0, 360, Scalar(0,255,255), 1, CV_AA);  //注意这里需要乘以0.5

最后，目前opencv3中支持三种椭圆拟合方式：fitEllipse，fitEllipseAMS，fitEllipseDirect，其中第三种来自《 Direct least squares fitting of ellipsees, Fitzgibbon, Pilu and Fischer in Fitzgibbon, A.W., Pilu, M., and Fischer R.B.,Proc. of the 13th Internation Conference on Pattern Recognition, pp 253–257, Vienna, 1996》

可以访问https://docs.opencv.org/trunk/de/dc7/fitellipse_8cpp-example.html

来看官方的椭圆拟合事例。