Triangulates set of corresponding 2D-2D points correspondances with known respective camera poses based on opengv. UUID: bb7dac37-499a-4bc4-9b57-3e010a94ed30 More...

#include <Triangulation.h>

Inheritance diagram for SolAR::MODULES::OPENGV::Triangulation:

[legend]

Collaboration diagram for SolAR::MODULES::OPENGV::Triangulation:

[legend]

Public Member Functions
	Triangulation ()
	Triangulation constructor.

	~Triangulation ()
	Triangulation 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...

double	triangulate (const std::vector< datastructure::Point2Df > &pt2d_1, const std::vector< datastructure::Point2Df > &pt2d_2, const std::vector< datastructure::DescriptorMatch > &matches, const std::pair< unsigned int, unsigned int > &working_views, const datastructure::Transform3Df &poseView1, const datastructure::Transform3Df &poseView2, std::vector< SRef< datastructure::CloudPoint > > &pcloud) override
	triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters). More...

double	triangulate (const std::vector< datastructure::Keypoint > &keypointsView1, const std::vector< datastructure::Keypoint > &keypointsView2, const std::vector< datastructure::DescriptorMatch > &matches, const std::pair< unsigned int, unsigned int > &working_views, const datastructure::Transform3Df &poseView1, const datastructure::Transform3Df &poseView2, std::vector< SRef< datastructure::CloudPoint > > &pcloud) override
	triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters). More...

double	triangulate (const std::vector< datastructure::Keypoint > &keypointsView1, const std::vector< datastructure::Keypoint > &keypointsView2, const SRef< datastructure::DescriptorBuffer > &descriptor1, const SRef< datastructure::DescriptorBuffer > &descriptor2, const std::vector< datastructure::DescriptorMatch > &matches, const std::pair< unsigned int, unsigned int > &working_views, const datastructure::Transform3Df &poseView1, const datastructure::Transform3Df &poseView2, std::vector< SRef< datastructure::CloudPoint > > &pcloud) override
	triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters). More...

double	triangulate (SRef< SolAR::datastructure::Frame > frame1, SRef< SolAR::datastructure::Frame > frame2, const std::vector< SolAR::datastructure::DescriptorMatch > &matches, const std::pair< uint32_t, uint32_t > &working_views, std::vector< SRef< SolAR::datastructure::CloudPoint > > &pcloud, const bool &onlyDepth=false) override
	calculating 3D cloud points by triangulating pairs of matched features or using depth information of keypoints. More...

void	unloadComponent () override final

Detailed Description

Triangulates set of corresponding 2D-2D points correspondances with known respective camera poses based on opengv. UUID: bb7dac37-499a-4bc4-9b57-3e010a94ed30

Member Function Documentation

◆ setCameraParameters()

void SolAR::MODULES::OPENGV::Triangulation::setCameraParameters	(	const datastructure::CamCalibration &	intrinsicParams,
		const datastructure::CamDistortion &	distorsionParams
	)

override

this method is used to set intrinsic parameters and distorsion of the camera

Parameters

[in]	Camera	calibration matrix parameters.
[in]	Camera	distorsion parameters.

◆ triangulate() [1/4]

double SolAR::MODULES::OPENGV::Triangulation::triangulate	(	const std::vector< datastructure::Keypoint > &	keypointsView1,
		const std::vector< datastructure::Keypoint > &	keypointsView2,
		const SRef< datastructure::DescriptorBuffer > &	descriptor1,
		const SRef< datastructure::DescriptorBuffer > &	descriptor2,
		const std::vector< datastructure::DescriptorMatch > &	matches,
		const std::pair< unsigned int, unsigned int > &	working_views,
		const datastructure::Transform3Df &	poseView1,
		const datastructure::Transform3Df &	poseView2,
		std::vector< SRef< datastructure::CloudPoint > > &	pcloud
	)

override

triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters).

Parameters

[in]	pointsView1,set	of keypoints seen in view_1.
[in]	pointsView2,set	of keypoints seen in view_2.
[in]	descriptor1,set	of descriptors in view_1.
[in]	descriptor2,set	of descriptors in view_2.
[in]	matches,the	matches between the keypoints of the view1 and the keypoints of the view 2.
[in]	working_views,a	pair representing the id of the two views
[in]	poseView1,Camera	pose in the world coordinates system of the view_1 expressed as a Transform3D.
[in]	poseView2,Camera	pose in the world coordinates system of the view_2 expressed as a Transform3D..
[out]	pcloud,Set	of triangulated 3d_points.

Returns: the mean re-projection error (mean distance in pixels between the original 2D points and the projection of the reconstructed 3D points)

◆ triangulate() [2/4]

double SolAR::MODULES::OPENGV::Triangulation::triangulate	(	const std::vector< datastructure::Keypoint > &	keypointsView1,
		const std::vector< datastructure::Keypoint > &	keypointsView2,
		const std::vector< datastructure::DescriptorMatch > &	matches,
		const std::pair< unsigned int, unsigned int > &	working_views,
		const datastructure::Transform3Df &	poseView1,
		const datastructure::Transform3Df &	poseView2,
		std::vector< SRef< datastructure::CloudPoint > > &	pcloud
	)

override

triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters).

Parameters

[in]	pointsView1,set	of keypoints seen in view_1.
[in]	pointsView2,set	of keypoints seen in view_2.
[in]	matches,the	matches between the keypoints of the view1 and the keypoints of the view 2.
[in]	working_views,a	pair representing the id of the two views
[in]	poseView1,Camera	pose in the world coordinates system of the view_1 expressed as a Transform3D.
[in]	poseView2,Camera	pose in the world coordinates system of the view_2 expressed as a Transform3D..
[out]	pcloud,Set	of triangulated 3d_points.

Returns: the mean re-projection error (mean distance in pixels between the original 2D points and the projection of the reconstructed 3D points)

◆ triangulate() [3/4]

double SolAR::MODULES::OPENGV::Triangulation::triangulate	(	const std::vector< datastructure::Point2Df > &	pt2d_1,
		const std::vector< datastructure::Point2Df > &	pt2d_2,
		const std::vector< datastructure::DescriptorMatch > &	matches,
		const std::pair< unsigned int, unsigned int > &	working_views,
		const datastructure::Transform3Df &	poseView1,
		const datastructure::Transform3Df &	poseView2,
		std::vector< SRef< datastructure::CloudPoint > > &	pcloud
	)

override

triangulate pairs of points 2d captured from two views with differents poses (with respect to the camera instrinsic parameters).

Parameters

[in]	pointsView1,set	of 2D points seen in view_1.
[in]	pointsView2,set	of 2D points seen in view_2.
[in]	matches,the	matches between the keypoints of the view1 and the keypoints of the view 2.
[in]	working_views,a	pair representing the id of the two views
[in]	poseView1,camera	pose in the world coordinates system of the view_1 expressed as a Transform3D.
[in]	poseView2,camera	pose in the world coordinates system of the view_2 expressed as a Transform3D..
[out]	pcloud,set	of triangulated 3d_points.

Returns: the mean re-projection error (mean distance in pixels between the original 2D points and the projection of the reconstructed 3D points)

◆ triangulate() [4/4]

double SolAR::MODULES::OPENGV::Triangulation::triangulate	(	SRef< SolAR::datastructure::Frame >	frame1,
		SRef< SolAR::datastructure::Frame >	frame2,
		const std::vector< SolAR::datastructure::DescriptorMatch > &	matches,
		const std::pair< uint32_t, uint32_t > &	working_views,
		std::vector< SRef< SolAR::datastructure::CloudPoint > > &	pcloud,
		const bool &	onlyDepth = `false`
	)

override

calculating 3D cloud points by triangulating pairs of matched features or using depth information of keypoints.

Parameters

[in]	frame1	the first frame.
[in]	frame2	the second frame.
[in]	matches	the matches between these two frames.
[in]	working_views	a pair representing the id of the two views
[out]	pcloud	Set of triangulated 3d_points.
[in]	onlyDepth	if it is true, using only depth information of keypoints for computing 3D cloud points.

Returns: the mean re-projection error (mean distance in pixels between the original 2D points and the projection of the reconstructed 3D points)

The documentation for this class was generated from the following file:

SolARModuleOpenGV/interfaces/Triangulation.h

Public Member Functions

Detailed Description

Member Function Documentation

◆ setCameraParameters()

◆ triangulate() [1/4]

◆ triangulate() [2/4]

◆ triangulate() [3/4]

◆ triangulate() [4/4]