Finds the camera pose based on a 2D-2D points correspondences between two images. UUID: 52babb5e-9d33-11e8-98d0-529269fb1459
More...
#include <SolARPoseFinderFrom2D2DOpencv.h>
Public Member Functions | |
| SolARPoseFinderFrom2D2DOpencv () | |
| SolARPoseFinderFrom2D2DOpencv constructor;. | |
| ~SolARPoseFinderFrom2D2DOpencv () override | |
| SolARPoseFinderFrom2D2DOpencv destructor;. | |
| void | setCameraParameters (const datastructure::CamCalibration &intrinsicParams, const datastructure::CamDistortion &distorsionParams) override |
| this method is used to set intrinsic parameters and distorsion of the camera More... | |
| FrameworkReturnCode | estimate (const std::vector< datastructure::Point2Df > &matchedPointsView1, const std::vector< datastructure::Point2Df > &matchedPointsView2, const datastructure::Transform3Df &poseView1, datastructure::Transform3Df &poseView2, std::vector< datastructure::DescriptorMatch > &inlierMatches) override |
| Estimates camera pose from a set of 2D points of the first image which match with a set of 2D points of the second image. More... | |
| FrameworkReturnCode | estimate (const std::vector< datastructure::Keypoint > &matchedPointsView1, const std::vector< datastructure::Keypoint > &matchedPointsView2, const datastructure::Transform3Df &poseView1, datastructure::Transform3Df &poseView2, std::vector< datastructure::DescriptorMatch > &inlierMatches) override |
| Estimates camera pose from a set of keypoints of the first image which match with a set of keypoints of the second image. More... | |
| void | unloadComponent () override final |
Detailed Description
Finds the camera pose based on a 2D-2D points correspondences between two images. UUID: 52babb5e-9d33-11e8-98d0-529269fb1459
This component first compute the essential Matrix between the two images, and then estimate the pose of the second camera (position and orientation of the camera defined in the world coordinate system) knowing the pose of the first camera. To do so, the component needs to know the intrinsic parameter of the camera. During the essential matrix estimation, the OpenCV method filters the matches based on a RANSAC approach by removing those that are to distant to the epipolar line. The estimation of the pose of the second camera is based on the recoverPose method of OpenCV which decomposes the essential matrix and selects among the four solutions the one that have the most important number of 3D points in front of the camera (cheirality check).
- Properties
outlierDistanceRatio here we are using a RANSAC method to remove outlier.
This attribute is the ratio between the maximum distance in pixels between source points and the maximum distance in pixels to the epipolar line for which point is considered as a outlier.
The higher is this ratio, the more you will keep inliers to estimate your 2D transform, but the less this estimation will be correct.
By default, this value is set to the one proposed by [Snavely07 4.1]
type: float; range : [0..MAX FLOAT]; default: 0.006f outlierDistanceRatio the probability that the algorithm produces a useful result type: float; range : [0..1]; default: 0.99f
Member Function Documentation
◆ estimate() [1/2]
|
override |
Estimates camera pose from a set of keypoints of the first image which match with a set of keypoints of the second image.
- Parameters
-
[in] pointsView1,Set of keypoints seen in view 1. [in] pointsView2,Set of keypoints seen in view 2 and matching with the 2D points of the view 1. [in] poseView1,Camera pose (3D transform of the camera of the view1 defined in world corrdinate system). [out] poseView2,Camera pose (3D transform of the camera of the view2 defined in world corrdinate system). [in|out] inlierMatches, a vector of matches that will be used for the pose estimation. This vector wll be updates as some input matches will be considered as outliers. If this vector is empty, we consider that the ith point of pointsView1 matches with the ith point of pointsView2.
◆ estimate() [2/2]
|
override |
Estimates camera pose from a set of 2D points of the first image which match with a set of 2D points of the second image.
- Parameters
-
[in] pointsView1,Set of 2D points seen in view 1. [in] pointsView2,Set of 2D points seen in view 2 and matching with the 2D points of the view 1. [in] poseView1,Camera pose (3D transform of the camera of the view1 defined in world corrdinate system). [out] poseView2,Camera pose (3D transform of the camera of the view2 defined in world corrdinate system). [in|out] inlierMatches, a vector of matches that will be used for the pose estimation. This vector wll be updates as some input matches will be considered as outliers. If this vector is empty, we consider that the ith point of pointsView1 matches with the ith point of pointsView2.
◆ setCameraParameters()
|
override |
this method is used to set intrinsic parameters and distorsion of the camera
- Parameters
-
[in] Camera calibration matrix parameters. [in] Camera distorsion parameters.
The documentation for this class was generated from the following file:
- SolARModuleOpenCV/interfaces/SolARPoseFinderFrom2D2DOpencv.h
