CorrespondenceEstimationBackprojection computes correspondences as points in the target cloud which have minimum More...

Public Member Functions
	NDEVRCorrespondenceEstimation ()
	Empty constructor.
virtual	~NDEVRCorrespondenceEstimation ()=default
	Empty destructor.
fltp04	calcDistance (uint04 a, uint04 b) const
	Calculate the weighted distance between two points, factoring in normal alignment.
void	determineCorrespondencesBackProjection (uint04 offset, pcl::Correspondences &correspondences, fltp04 max_distance)
	Determine correspondences using back-projection along normals.
void	determineCorrespondencesMinDistance (uint04 offset, pcl::Correspondences &correspondences, fltp04 max_distance)
	Determine correspondences using minimum distance search.
unsigned int	getKSearch () const
	Get the number of nearest neighbours considered in the target point cloud for computing correspondences.
NormalsConstPtr	getSourceNormals () const
	Get the normals of the source point cloud.
NormalsConstPtr	getTargetNormals () const
	Get the normals of the target point cloud.
LineSegment< 3, fltp04 >	lineSegment (uint04 index) const
	Returns a line segment along the normal at the given index, expanded by m_normal_expansion_factor.
MaxHeap< fltp04, uint04 >	radiusKSearch (uint04 tgt_idx, fltp04 max_distance_squared, uint04 k=0) const
	Find k closest points within a maximum distance.
MaxHeap< fltp04, uint04 >	radiusNormalKSearch (uint04 tgt_idx, fltp04 max_distance_squared, uint04 k=0) const
	Find k closest points using normal-expanded line segment search within a max distance.
bool	radiusNormalSearch (uint04 tgt_idx, uint04 &index, fltp04 &k_sqr_distances) const
	Find the closest point in the tree using normal-expanded line segment search.
bool	radiusSearch (uint04 tgt_idx, uint04 &index, fltp04 &k_sqr_distances) const
	Find the closest point in the tree to the target point.
bool	requiresSourceNormals () const
	See if this rejector requires source normals.
bool	requiresTargetNormals () const
	See if this rejector requires target normals.
void	setInputTarget (PointCloudSourcePtr input)
	Set the input target point cloud.
void	setKSearch (unsigned int k)
	Set the number of nearest neighbours to be considered in the target point cloud.
void	setSourceNormals (const NormalsConstPtr &normals)
	Set the normals computed on the source point cloud.
void	setSourceNormals (pcl::PCLPointCloud2::ConstPtr cloud2)
	Blob method for setting the source normals.
void	setTargetNormals (const NormalsConstPtr &normals)
	Set the normals computed on the target point cloud.
void	setTargetNormals (pcl::PCLPointCloud2::ConstPtr cloud2)
	Method for setting the target normals.
void	setTree (const RTree< 3, fltp04 > tree, const Buffer< Vertex< 3, fltp04 > > params, const Buffer< Vertex< 3, fltp04 > > *tree_sorted_params)
	Set the spatial search tree and associated point data.
Vertex< 3, fltp04 >	vertex (uint04 index) const
	Determine the correspondences between input and target cloud.

Detailed Description

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>
class NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >

CorrespondenceEstimationBackprojection computes correspondences as points in the target cloud which have minimum

Author: Suat Gedikli

Definition at line 20 of file NDEVRCorrespondenceEstimation.h.

Constructor & Destructor Documentation

◆ NDEVRCorrespondenceEstimation()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::NDEVRCorrespondenceEstimation ( )

inline

Empty constructor.

Note: Sets the number of neighbors to be considered in the target point cloud (k_) to 10.

Definition at line 49 of file NDEVRCorrespondenceEstimation.h.

Member Function Documentation

◆ calcDistance()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

fltp04 NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::calcDistance	(	uint04	a,
		uint04	b ) const

inline

Calculate the weighted distance between two points, factoring in normal alignment.

Parameters

[in]	a	index of the first point
[in]	b	index of the second point

Returns: the weighted distance value

Definition at line 236 of file NDEVRCorrespondenceEstimation.h.

References vertex().

Referenced by determineCorrespondencesBackProjection().

◆ determineCorrespondencesBackProjection()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::determineCorrespondencesBackProjection	(	uint04	offset,
		pcl::Correspondences &	correspondences,
		fltp04	max_distance )

inline

Determine correspondences using back-projection along normals.

Parameters

[in]	offset	the starting index offset into the input cloud
[out]	correspondences	the found correspondences
[in]	max_distance	the maximum allowed distance for a correspondence

Definition at line 280 of file NDEVRCorrespondenceEstimation.h.

References calcDistance(), radiusNormalKSearch(), and radiusNormalSearch().

◆ determineCorrespondencesMinDistance()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::determineCorrespondencesMinDistance	(	uint04	offset,
		pcl::Correspondences &	correspondences,
		fltp04	max_distance )

inline

Determine correspondences using minimum distance search.

Parameters

[in]	offset	the starting index offset into the input cloud
[out]	correspondences	the found correspondences
[in]	max_distance	the maximum allowed distance for a correspondence

Definition at line 252 of file NDEVRCorrespondenceEstimation.h.

References radiusSearch().

◆ getKSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

unsigned int NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::getKSearch ( ) const

inline

Get the number of nearest neighbours considered in the target point cloud for computing correspondences.

By default we use k = 10 nearest neighbors.

Definition at line 348 of file NDEVRCorrespondenceEstimation.h.

◆ lineSegment()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

LineSegment< 3, fltp04 > NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::lineSegment ( uint04 index ) const

inline

Returns a line segment along the normal at the given index, expanded by m_normal_expansion_factor.

Parameters

[in] index the point index

Returns: the line segment along the normal direction

Definition at line 153 of file NDEVRCorrespondenceEstimation.h.

Referenced by radiusNormalKSearch(), and radiusNormalSearch().

◆ radiusKSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

MaxHeap< fltp04, uint04 > NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::radiusKSearch	(	uint04	tgt_idx,
		fltp04	max_distance_squared,
		uint04	k = 0 ) const

inline

Find k closest points within a maximum distance.

Parameters

[in]	tgt_idx	the target point index
[in]	max_distance_squared	the maximum squared search distance
[in]	k	the number of neighbors to find

Returns: a max-heap of (distance, index) pairs

Definition at line 201 of file NDEVRCorrespondenceEstimation.h.

References cast(), and vertex().

◆ radiusNormalKSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

MaxHeap< fltp04, uint04 > NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::radiusNormalKSearch	(	uint04	tgt_idx,
		fltp04	max_distance_squared,
		uint04	k = 0 ) const

inline

Find k closest points using normal-expanded line segment search within a max distance.

Parameters

[in]	tgt_idx	the target point index
[in]	max_distance_squared	the maximum squared search distance
[in]	k	the number of neighbors to find

Returns: a max-heap of (distance, index) pairs

Definition at line 219 of file NDEVRCorrespondenceEstimation.h.

References cast(), and lineSegment().

Referenced by determineCorrespondencesBackProjection().

◆ radiusNormalSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

bool NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::radiusNormalSearch	(	uint04	tgt_idx,
		uint04 &	index,
		fltp04 &	k_sqr_distances ) const

inline

Find the closest point in the tree using normal-expanded line segment search.

Parameters

[in]	tgt_idx	the target point index
[out]	index	the found closest point index
[out]	k_sqr_distances	the squared distance to the closest point

Returns: true if a neighbor was found

Definition at line 186 of file NDEVRCorrespondenceEstimation.h.

References IsInvalid(), and lineSegment().

Referenced by determineCorrespondencesBackProjection().

◆ radiusSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

bool NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::radiusSearch	(	uint04	tgt_idx,
		uint04 &	index,
		fltp04 &	k_sqr_distances ) const

inline

Find the closest point in the tree to the target point.

Parameters

[in]	tgt_idx	the target point index
[out]	index	the found closest point index
[out]	k_sqr_distances	the squared distance to the closest point

Returns: true if a neighbor was found

Definition at line 171 of file NDEVRCorrespondenceEstimation.h.

References IsInvalid(), and vertex().

Referenced by determineCorrespondencesMinDistance().

◆ setInputTarget()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::setInputTarget ( PointCloudSourcePtr input )

inline

Set the input target point cloud.

Parameters

[in] input the target point cloud

Definition at line 107 of file NDEVRCorrespondenceEstimation.h.

◆ setKSearch()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::setKSearch ( unsigned int k )

inline

Set the number of nearest neighbours to be considered in the target point cloud.

By default, we use k = 10 nearest neighbors.

Parameters

[in] k the number of nearest neighbours to be considered

Definition at line 340 of file NDEVRCorrespondenceEstimation.h.

◆ setSourceNormals()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::setSourceNormals ( const NormalsConstPtr & normals )

inline

Set the normals computed on the source point cloud.

Parameters

[in] normals the normals computed for the source cloud

Definition at line 72 of file NDEVRCorrespondenceEstimation.h.

Referenced by setSourceNormals().

◆ setTargetNormals()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::setTargetNormals ( const NormalsConstPtr & normals )

inline

Set the normals computed on the target point cloud.

Parameters

[in] normals the normals computed for the target cloud

Definition at line 87 of file NDEVRCorrespondenceEstimation.h.

Referenced by setTargetNormals().

◆ setTree()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

void NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::setTree	(	const RTree< 3, fltp04 > *	tree,
		const Buffer< Vertex< 3, fltp04 > > *	params,
		const Buffer< Vertex< 3, fltp04 > > *	tree_sorted_params )

inline

Set the spatial search tree and associated point data.

Parameters

[in]	tree	the R-tree spatial index
[in]	params	the point buffer
[in]	tree_sorted_params	the pre-sorted point buffer for tree queries

Definition at line 63 of file NDEVRCorrespondenceEstimation.h.

◆ vertex()

template<typename PointSource, typename PointTarget, typename NormalT, typename Scalar = float>

Vertex< 3, fltp04 > NDEVRCorrespondenceEstimation< PointSource, PointTarget, NormalT, Scalar >::vertex ( uint04 index ) const

inline

Determine the correspondences between input and target cloud.

Parameters

[out]	correspondences	the found correspondences (index of query point, index of target point, distance)
[in]	max_distance	maximum distance between the normal on the source point cloud and the corresponding point in the target point cloud

Returns the 3D vertex at the given index in the input cloud.

Parameters

[in] index the point index

Returns: the 3D vertex

Definition at line 144 of file NDEVRCorrespondenceEstimation.h.

Referenced by calcDistance(), radiusKSearch(), and radiusSearch().

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

LocationRegistration/Headers/NDEVRCorrespondenceEstimation.h

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ NDEVRCorrespondenceEstimation()

Member Function Documentation

◆ calcDistance()

◆ determineCorrespondencesBackProjection()

◆ determineCorrespondencesMinDistance()

◆ getKSearch()

◆ lineSegment()

◆ radiusKSearch()

◆ radiusNormalKSearch()

◆ radiusNormalSearch()

◆ radiusSearch()

◆ setInputTarget()

◆ setKSearch()

◆ setSourceNormals()

◆ setTargetNormals()

◆ setTree()

◆ vertex()