参数详解：
srcPoints 源平面中点的坐标矩阵，可以是CV_32FC2类型，也可以是vector<Point2f>类型
dstPoints 目标平面中点的坐标矩阵，可以是CV_32FC2类型，也可以是vector<Point2f>类型
method 计算单应矩阵所使用的方法。不同的方法对应不同的参数，具体如下：
0 - 利用所有点的常规方法
RANSAC - RANSAC-基于RANSAC的鲁棒算法
LMEDS - 最小中值鲁棒算法
RHO - PROSAC-基于PROSAC的鲁棒算法
ransacReprojThreshold
将点对视为内点的最大允许重投影错误阈值（仅用于RANSAC和RHO方法）。如果
则点被认为是个外点（即错误匹配点对）。若srcPoints和dstPoints是以像素为单位的，则该参数通常设置在1到10的范围内。
mask
可选输出掩码矩阵，通常由鲁棒算法（RANSAC或LMEDS）设置。 请注意，输入掩码矩阵是不需要设置的。
maxIters RANSAC 算法的最大迭代次数，默认值为2000。
confidence 可信度值，取值范围为0到1.
首先定义两个vector保存对应的4对点
//图片映射矩阵把不同角度的图片矫正
void findHomographyText(){
// Read source image.
Mat src = imread("F:\\视觉\\opencv\\pic\\1.png");
// Four corners of the book in source image
vector<Point2f> pts_src;
pts_src.push_back(Point2f(0, 0));
pts_src.push_back(Point2f(src.cols, 0));
pts_src.push_back(Point2f(src.cols, src.rows));
pts_src.push_back(Point2f(0, src.rows));
// Four corners of the book in destination image.
vector<Point2f> pts_dst;
pts_dst.push_back(Point2f(0, 0));
pts_dst.push_back(Point2f(src.cols/4, 0));
pts_dst.push_back(Point2f(src.cols/3, src.rows));
pts_dst.push_back(Point2f(0, src.rows/2));
// Calculate Homography
Mat h = findHomography(pts_src, pts_dst);
// Output image
Mat im_out;
// Warp source image to destination based on homography
warpPerspective(src, im_out, h, src.size());
// Display images
imshow("Source Image", src);
imshow("Warped Source Image", im_out);
waitKey(0);
结果如下图所示对图像进行拉伸
1,相求H
vector<Point2f> pts_src;
pts_src.push_back(Point2f(0, 0));
pts_src.push_back(Point2f(src.cols, 0));
pts_src.push_back(Point2f(src.cols, src.rows));
pts_src.push_back(Point2f(0, src.rows));
// Four corners of the book in destination image.
vector<Point2f> pts_dst;
pts_dst.push_back(Point2f(0, 0));
pts_dst.push_back(Point2f(src.cols/4, 0));
pts_dst.push_back(Point2f(src.cols/3, src.rows));
pts_dst.push_back(Point2f(0, src.rows/2));
// Calculate Homography
Mat h = findHomography(pts_src, pts_dst);
通过H求对应的图像（映射到输出图片上）
warpPerspective(src, im_out, h, src.size());
warpPerspective：通过H求取
im_out输出值介绍完两个主要的函数下面开始对图像进行识别和标记
2，SURF对图像的识别和标记
1，开发思路
（1）使用SIFT或者SURF进行角点检测，获取两个图像的的角点集合
（2）根据两个集合，使用特征点匹配，匹配类似的点 FlannBasedMatcher
（3）过滤特征点对。
（4）通过特征点对，求出H值
（5）画出特征区域
代码实现：
1，使用SIFT或者SURF进行角点检测，获取两个图像的的角点集合
src = imread("F:\\视觉\\opencv\\pic\\11.png");//读图片
src3 = imread("F:\\视觉\\opencv\\pic\\5.png");//读图片
int minHessian = 400;
cvtColor(src, src, COLOR_BGR2GRAY);
cvtColor(src3, src3, COLOR_BGR2GRAY);
Ptr<SIFT> detector = SIFT::create(minHessian);
vector<KeyPoint> keypoints_obj;//图片1特征点
vector<KeyPoint> keypoints_scene;//图片2特征点
Mat descriptor_obj, descriptor_scene;
//找出特征点存到keypoints_obj与keypoints_scene点集中
detector->detectAndCompute(src, Mat(), keypoints_obj, descriptor_obj);
detector->detectAndCompute(src3, Mat(), keypoints_scene, descriptor_scene);
// matching 找到特征集合
FlannBasedMatcher matcher;
vector<DMatch> matches;
matcher.match(descriptor_obj, descriptor_scene, matches);
2，过滤相似度高的图像
// find good matched points
double minDist = 1000;
double maxDist = 0;
for (int i = 0; i < descriptor_obj.rows; i++) {
double dist = matches[i].distance;
if (dist > maxDist) {
maxDist = dist;
if (dist < minDist) {
minDist = dist;
printf("max distance : %f\n", maxDist);
printf("min distance : %f\n", minDist);
vector<DMatch> goodMatches;
//过滤相同的点
for (int i = 0; i < descriptor_obj.rows; i++) {
double dist = matches[i].distance;//相识度
printf("distance : %f\n", dist);
if (dist < max(3 * minDist, 0.2)) {
goodMatches.push_back(matches[i]);
3，求出H
vector<Point2f> obj;
vector<Point2f> objInScene;
for (size_t t = 0; t < goodMatches.size(); t++) {
//把DMatch转成坐标 Point2f
obj.push_back(keypoints_obj[goodMatches[t].queryIdx].pt);
objInScene.push_back(keypoints_scene[goodMatches[t].trainIdx].pt);
//用来求取“射影变换”的H转制矩阵函数 X'=H X ，并使用RANSAC消除一些出错的点
Mat H = findHomography(obj, objInScene, RANSAC);
4,使用H求出映射到大图的点
vector<Point2f> obj_corners(4);
vector<Point2f> scene_corners(4);
obj_corners[0] = Point(0, 0);
obj_corners[1] = Point(src.cols, 0);
obj_corners[2] = Point(src.cols, src.rows);
obj_corners[3] = Point(0, src.rows);
//透视变换(把斜的图片扶正)
cout << H << endl;
perspectiveTransform(obj_corners, scene_corners, H);
5，在原图上画线段
Mat dst;
cvtColor(src3, dst, COLOR_GRAY2BGR);
line(dst, scene_corners[0], scene_corners[1], Scalar(0, 0, 255), 2, 8, 0);
line(dst, scene_corners[1], scene_corners[2], Scalar(0, 0, 255), 2, 8, 0);
line(dst, scene_corners[2], scene_corners[3], Scalar(0, 0, 255), 2, 8, 0);
line(dst, scene_corners[3], scene_corners[0], Scalar(0, 0, 255), 2, 8, 0);
imshow("Draw object", dst);
谢谢，如果觉得可以请点个赞！转发请付链接。。。。
原文链接：https://blog.csdn.net/fengyeer20120/article/details/87798638
https://www.cnblogs.com/wangguchangqing/p/4645805.html
// find good matched pointsdouble minDist = 1000;double maxDist = 0;
for (int i = 0; i < descriptor_obj.rows; i++) {double dist = matches[i].distance;if (dist > maxDist) {maxDist = dist;}if (dist < minDist) {minDist = dist;}}printf("max distance : %f\n", maxDist);printf("min distance : %f\n", minDist);
vector<DMatch> goodMatches;//过滤相同的点for (int i = 0; i < descriptor_obj.rows; i++) {double dist = matches[i].distance;//相识度printf("distance : %f\n", dist);if (dist < max(3 * minDist, 0.2)) {goodMatches.push_back(matches[i]);}}
