写一个orb特征点匹配的c++代码，要求如下：1 将shi-tomasi算法加入 2 加入GMS算法和ransac算法去除误匹配

由于orb特征点匹配的代码较为复杂，下面只给出主要的代码框架和一些关键点的注释。

首先是包含头文件和命名空间声明：

#include <opencv2/opencv.hpp>
#include <opencv2/features2d/features2d.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#include <vector>

using namespace cv;
using namespace std;

然后是定义主函数：

int main(int argc, char** argv)
{
    // 读取图像
    Mat img1 = imread(argv[1], CV_LOAD_IMAGE_GRAYSCALE);
    Mat img2 = imread(argv[2], CV_LOAD_IMAGE_GRAYSCALE);

    // 定义ORB特征点检测器
    Ptr<ORB> orb = ORB::create();

    // 定义Shi-Tomasi角点检测器
    int max_corners = 500;
    double quality_level = 0.01;
    double min_distance = 10;
    int block_size = 3;
    bool use_harris_detector = false;
    double k = 0.04;
    vector<Point2f> corners1, corners2;
    goodFeaturesToTrack(img1, corners1, max_corners, quality_level, min_distance, Mat(), block_size, use_harris_detector, k);
    goodFeaturesToTrack(img2, corners2, max_corners, quality_level, min_distance, Mat(), block_size, use_harris_detector, k);

    // 提取ORB特征点
    vector<KeyPoint> keypoints1, keypoints2;
    orb->detect(img1, keypoints1);
    orb->detect(img2, keypoints2);

    // 计算ORB描述子
    Mat descriptors1, descriptors2;
    orb->compute(img1, keypoints1, descriptors1);
    orb->compute(img2, keypoints2, descriptors2);

    // 匹配ORB特征点
    BFMatcher matcher(NORM_HAMMING);
    vector<DMatch> matches;
    matcher.match(descriptors1, descriptors2, matches);

    // 前n个匹配特征点
    int n = matches.size();
    vector<Point2f> points1, points2;
    for (int i = 0; i < n; i++)
    {
        points1.push_back(keypoints1[matches[i].queryIdx].pt);
        points2.push_back(keypoints2[matches[i].trainIdx].pt);
    }

    // GMS算法去除误匹配
    vector<bool> inliers;
    vector<DMatch> gms_matches;
    bool result = GMS(points1, img1.size(), points2, img2.size(), matches, gms_matches, inliers);

    // RANSAC算法去除误匹配
    vector<Point2f> inlier_points1, inlier_points2;
    for (int i = 0; i < n; i++)
    {
        if (inliers[i])
        {
            inlier_points1.push_back(points1[i]);
            inlier_points2.push_back(points2[i]);
        }
    }
    Mat H = findHomography(inlier_points1, inlier_points2, RANSAC);

    // 显示匹配结果
    Mat img_matches;
    drawMatches(img1, keypoints1, img2, keypoints2, gms_matches, img_matches);
    imshow("Matches", img_matches);
    waitKey(0);
    return 0;
}

其中，Shi-Tomasi角点检测器的参数需要根据图像具体情况进行调整，GMS算法和RANSAC算法的实现可以使用OpenCV提供的函数。