已知内外参，用Opencv c++实现一个图像2d点到3d点的转换函数

可以使用Opencv中的`cv::projectPoints`函数实现2D点到3D点的转换。该函数的使用方式如下：

void cv::projectPoints(InputArray objectPoints, InputArray rvec, InputArray tvec, InputArray cameraMatrix, InputArray distCoeffs, OutputArray imagePoints);

其中，`objectPoints`为输入的2D点坐标，`rvec`和`tvec`为相机的旋转向量和平移向量，`cameraMatrix`为相机内参矩阵，`distCoeffs`为相机的畸变参数，`imagePoints`为输出的3D点坐标。

具体使用方法如下：

// 读入相机内参矩阵和畸变参数
cv::Mat cameraMatrix = (cv::Mat_<double>(3, 3) << fx, 0, cx, 0, fy, cy, 0, 0, 1);
cv::Mat distCoeffs = (cv::Mat_<double>(1, 5) << k1, k2, p1, p2, k3);

// 定义旋转向量和平移向量
cv::Mat rvec, tvec;

// 读入2D点坐标
std::vector<cv::Point2f> imagePoints; // 2D点坐标
for (int i = 0; i < num_points; i++) {
    imagePoints.push_back(cv::Point2f(x[i], y[i]));
}

// 2D点到3D点转换
std::vector<cv::Point3f> objectPoints; // 3D点坐标
cv::Mat imagePointsMat(imagePoints);
cv::Mat objectPointsMat;
cv::undistortPoints(imagePointsMat, objectPointsMat, cameraMatrix, distCoeffs);
cv::solvePnP(objectPointsMat, imagePointsMat, cameraMatrix, distCoeffs, rvec, tvec);

// 输出3D点坐标
std::vector<cv::Point3f> outputPoints;
cv::projectPoints(objectPointsMat, rvec, tvec, cameraMatrix, distCoeffs, outputPoints);

// 将3D点坐标输出到数组中
for (int i = 0; i < num_points; i++) {
    x3d[i] = outputPoints[i].x;
    y3d[i] = outputPoints[i].y;
    z3d[i] = outputPoints[i].z;
}