| namespace op | |
| { | |
| /** | |
| * 3D triangulation given known camera parameter matrices and based on linear DLT algorithm. | |
| * The returned cv::Mat is a 4x1 matrix, where the last coordinate is 1. | |
| */ | |
| double triangulate( | |
| cv::Mat& reconstructedPoint, const std::vector<cv::Mat>& cameraMatrices, | |
| const std::vector<cv::Point2d>& pointsOnEachCamera); | |
| /** | |
| * 3D triangulation given known camera parameter matrices and based on linear DLT algorithm with additional LMA | |
| * non-linear refinement. | |
| * The returned cv::Mat is a 4x1 matrix, where the last coordinate is 1. | |
| * Note: If Ceres is not enabled, the LMA refinement is skipped and this function is equivalent to triangulate(). | |
| */ | |
| double triangulateWithOptimization( | |
| cv::Mat& reconstructedPoint, const std::vector<cv::Mat>& cameraMatrices, | |
| const std::vector<cv::Point2d>& pointsOnEachCamera, const double reprojectionMaxAcceptable); | |
| } | |