CorrespondenceRejectorSampleConsensus implements a correspondence rejection using Random Sample Consensus to identify inliers (and reject outliers). More...

#include <pcl/registration/correspondence_rejection_sample_consensus.h>

Inheritance diagram for pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >:

Public Types
using	Ptr = shared_ptr<CorrespondenceRejectorSampleConsensus<PointT>>
using	ConstPtr = shared_ptr<const CorrespondenceRejectorSampleConsensus<PointT>>
Public Types inherited from pcl::registration::CorrespondenceRejector
using	Ptr = shared_ptr<CorrespondenceRejector>
using	ConstPtr = shared_ptr<const CorrespondenceRejector>

Public Member Functions
	CorrespondenceRejectorSampleConsensus ()
	Empty constructor.
	~CorrespondenceRejectorSampleConsensus () override=default
	Empty destructor.
void	getRemainingCorrespondences (const pcl::Correspondences &original_correspondences, pcl::Correspondences &remaining_correspondences) override
	Get a list of valid correspondences after rejection from the original set of correspondences.
virtual void	setInputSource (const PointCloudConstPtr &cloud)
	Provide a source point cloud dataset (must contain XYZ data!).
PointCloudConstPtr const	getInputSource ()
	Get a pointer to the input point cloud dataset target.
virtual void	setInputTarget (const PointCloudConstPtr &cloud)
	Provide a target point cloud dataset (must contain XYZ data!).
PointCloudConstPtr const	getInputTarget ()
	Get a pointer to the input point cloud dataset target.
bool	requiresSourcePoints () const override
	See if this rejector requires source points.
void	setSourcePoints (pcl::PCLPointCloud2::ConstPtr cloud2) override
	Blob method for setting the source cloud.
bool	requiresTargetPoints () const override
	See if this rejector requires a target cloud.
void	setTargetPoints (pcl::PCLPointCloud2::ConstPtr cloud2) override
	Method for setting the target cloud.
void	setInlierThreshold (double threshold)
	Set the maximum distance between corresponding points.
double	getInlierThreshold ()
	Get the maximum distance between corresponding points.
void	setMaximumIterations (int max_iterations)
	Set the maximum number of iterations.
int	getMaximumIterations ()
	Get the maximum number of iterations.
Eigen::Matrix4f	getBestTransformation ()
	Get the best transformation after RANSAC rejection.
void	setRefineModel (const bool refine)
	Specify whether the model should be refined internally using the variance of the inliers.
bool	getRefineModel () const
	Get the internal refine parameter value as set by the user using setRefineModel.
void	getInliersIndices (pcl::Indices &inlier_indices)
	Get the inlier indices found by the correspondence rejector.
void	setSaveInliers (bool s)
	Set whether to save inliers or not.
bool	getSaveInliers ()
	Get whether the rejector is configured to save inliers.
const std::string &	getClassName () const
	Get a string representation of the name of this class.
Public Member Functions inherited from pcl::registration::CorrespondenceRejector
	CorrespondenceRejector ()=default
	Empty constructor.
virtual	~CorrespondenceRejector ()=default
	Empty destructor.
virtual void	setInputCorrespondences (const CorrespondencesConstPtr &correspondences)
	Provide a pointer to the vector of the input correspondences.
CorrespondencesConstPtr	getInputCorrespondences ()
	Get a pointer to the vector of the input correspondences.
void	getCorrespondences (pcl::Correspondences &correspondences)
	Run correspondence rejection.
void	getRejectedQueryIndices (const pcl::Correspondences &correspondences, pcl::Indices &indices)
	Determine the indices of query points of correspondences that have been rejected, i.e., the difference between the input correspondences (set via setInputCorrespondences) and the given correspondence vector.
const std::string &	getClassName () const
	Get a string representation of the name of this class.
virtual bool	requiresSourceNormals () const
	See if this rejector requires source normals.
virtual void	setSourceNormals (pcl::PCLPointCloud2::ConstPtr)
	Abstract method for setting the source normals.
virtual bool	requiresTargetNormals () const
	See if this rejector requires target normals.
virtual void	setTargetNormals (pcl::PCLPointCloud2::ConstPtr)
	Abstract method for setting the target normals.

Public Attributes
CorrespondencesConstPtr	input_correspondences_
	The input correspondences.
std::string	rejection_name_
	The name of the rejection method.

Protected Member Functions
void	applyRejection (pcl::Correspondences &correspondences) override
	Apply the rejection algorithm.

Protected Attributes
double	inlier_threshold_ {0.05}
int	max_iterations_ {1000}
PointCloudConstPtr	input_
PointCloudPtr	input_transformed_
PointCloudConstPtr	target_
Eigen::Matrix4f	best_transformation_
bool	refine_ {false}
pcl::Indices	inlier_indices_
bool	save_inliers_ {false}
Protected Attributes inherited from pcl::registration::CorrespondenceRejector
std::string	rejection_name_ {}
	The name of the rejection method.
CorrespondencesConstPtr	input_correspondences_
	The input correspondences.

Detailed Description

template<typename PointT>
class pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >

CorrespondenceRejectorSampleConsensus implements a correspondence rejection using Random Sample Consensus to identify inliers (and reject outliers).

Author: Dirk Holz

Definition at line 54 of file correspondence_rejection_sample_consensus.h.

Member Typedef Documentation

◆ ConstPtr

template<typename PointT>

using pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::ConstPtr = shared_ptr<const CorrespondenceRejectorSampleConsensus<PointT>>

Definition at line 65 of file correspondence_rejection_sample_consensus.h.

◆ Ptr

template<typename PointT>

using pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::Ptr = shared_ptr<CorrespondenceRejectorSampleConsensus<PointT>>

Definition at line 64 of file correspondence_rejection_sample_consensus.h.

Constructor & Destructor Documentation

◆ CorrespondenceRejectorSampleConsensus()

template<typename PointT>

pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::CorrespondenceRejectorSampleConsensus ( )

inline

Empty constructor.

Sets the inlier threshold to 5cm (0.05m), and the maximum number of iterations to 1000.

Definition at line 70 of file correspondence_rejection_sample_consensus.h.

References input_, input_transformed_, rejection_name_, and target_.

◆ ~CorrespondenceRejectorSampleConsensus()

template<typename PointT>

pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::~CorrespondenceRejectorSampleConsensus ( )

overridedefault

Empty destructor.

Member Function Documentation

◆ applyRejection()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::applyRejection ( pcl::Correspondences & correspondences )

inlineoverrideprotectedvirtual

Apply the rejection algorithm.

Parameters

[out] correspondences the set of resultant correspondences.

Implements pcl::registration::CorrespondenceRejector.

Reimplemented in pcl::registration::CorrespondenceRejectorSampleConsensus2D< PointT >.

Definition at line 247 of file correspondence_rejection_sample_consensus.h.

References getRemainingCorrespondences(), and input_correspondences_.

◆ getBestTransformation()

template<typename PointT>

Eigen::Matrix4f pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getBestTransformation ( )

inline

Get the best transformation after RANSAC rejection.

Returns: The homogeneous 4x4 transformation yielding the largest number of inliers.

Definition at line 193 of file correspondence_rejection_sample_consensus.h.

References best_transformation_.

Referenced by pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), and pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences().

◆ getClassName()

template<typename PointT>

const std::string & pcl::registration::CorrespondenceRejector::getClassName ( ) const

inline

Get a string representation of the name of this class.

Definition at line 131 of file correspondence_rejection.h.

Referenced by pcl::registration::CorrespondenceRejectorSampleConsensus2D< PointT >::getRemainingCorrespondences(), and getRemainingCorrespondences().

◆ getInliersIndices()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getInliersIndices ( pcl::Indices & inlier_indices )

inline

Get the inlier indices found by the correspondence rejector.

This information is only saved if setSaveInliers(true) was called in advance.

Parameters

[out] inlier_indices Indices for the inliers

Definition at line 221 of file correspondence_rejection_sample_consensus.h.

References inlier_indices_.

◆ getInlierThreshold()

template<typename PointT>

double pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getInlierThreshold ( )

inline

Get the maximum distance between corresponding points.

Returns: Distance threshold in the same dimension as source and target data sets.

Definition at line 166 of file correspondence_rejection_sample_consensus.h.

References inlier_threshold_.

◆ getInputSource()

template<typename PointT>

PointCloudConstPtr const pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getInputSource ( )

inline

Get a pointer to the input point cloud dataset target.

Definition at line 98 of file correspondence_rejection_sample_consensus.h.

References input_.

◆ getInputTarget()

template<typename PointT>

PointCloudConstPtr const pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getInputTarget ( )

inline

Get a pointer to the input point cloud dataset target.

Definition at line 114 of file correspondence_rejection_sample_consensus.h.

References target_.

◆ getMaximumIterations()

template<typename PointT>

int pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getMaximumIterations ( )

inline

Get the maximum number of iterations.

Returns: max_iterations Maximum number if iterations to run

Definition at line 184 of file correspondence_rejection_sample_consensus.h.

References max_iterations_.

◆ getRefineModel()

template<typename PointT>

bool pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getRefineModel ( ) const

inline

Get the internal refine parameter value as set by the user using setRefineModel.

Definition at line 211 of file correspondence_rejection_sample_consensus.h.

References refine_.

◆ getRemainingCorrespondences()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getRemainingCorrespondences	(	const pcl::Correspondences &	original_correspondences,
		pcl::Correspondences &	remaining_correspondences )

inlineoverridevirtual

Get a list of valid correspondences after rejection from the original set of correspondences.

Parameters

[in]	original_correspondences	the set of initial correspondences given
[out]	remaining_correspondences	the resultant filtered set of remaining correspondences

Implements pcl::registration::CorrespondenceRejector.

Reimplemented in pcl::registration::CorrespondenceRejectorSampleConsensus2D< PointT >.

Definition at line 55 of file correspondence_rejection_sample_consensus.hpp.

References best_transformation_, pcl::RandomSampleConsensus< PointT >::computeModel(), getClassName(), pcl::SampleConsensus< T >::getInliers(), pcl::SampleConsensus< T >::getModelCoefficients(), inlier_indices_, inlier_threshold_, input_, max_iterations_, refine_, pcl::SampleConsensus< T >::refineModel(), save_inliers_, pcl::SampleConsensus< T >::setMaxIterations(), and target_.

Referenced by applyRejection(), pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), and pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences().

◆ getSaveInliers()

template<typename PointT>

bool pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::getSaveInliers ( )

inline

Get whether the rejector is configured to save inliers.

Definition at line 237 of file correspondence_rejection_sample_consensus.h.

References save_inliers_.

◆ requiresSourcePoints()

template<typename PointT>

bool pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::requiresSourcePoints ( ) const

inlineoverridevirtual

See if this rejector requires source points.

Reimplemented from pcl::registration::CorrespondenceRejector.

Definition at line 121 of file correspondence_rejection_sample_consensus.h.

◆ requiresTargetPoints()

template<typename PointT>

bool pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::requiresTargetPoints ( ) const

inlineoverridevirtual

See if this rejector requires a target cloud.

Reimplemented from pcl::registration::CorrespondenceRejector.

Definition at line 137 of file correspondence_rejection_sample_consensus.h.

◆ setInlierThreshold()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setInlierThreshold ( double threshold )

inline

Set the maximum distance between corresponding points.

Correspondences with distances below the threshold are considered as inliers.

Parameters

[in] threshold Distance threshold in the same dimension as source and target data sets.

Definition at line 157 of file correspondence_rejection_sample_consensus.h.

References inlier_threshold_.

Referenced by pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), and pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences().

◆ setInputSource()

template<typename PointT>

virtual void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setInputSource ( const PointCloudConstPtr & cloud )

inlinevirtual

Provide a source point cloud dataset (must contain XYZ data!).

Parameters

[in] cloud a cloud containing XYZ data

Definition at line 91 of file correspondence_rejection_sample_consensus.h.

References input_.

Referenced by pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences(), and setSourcePoints().

◆ setInputTarget()

template<typename PointT>

virtual void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setInputTarget ( const PointCloudConstPtr & cloud )

inlinevirtual

Provide a target point cloud dataset (must contain XYZ data!).

Parameters

[in] cloud a cloud containing XYZ data

Definition at line 107 of file correspondence_rejection_sample_consensus.h.

References target_.

Referenced by pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences(), and setTargetPoints().

◆ setMaximumIterations()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setMaximumIterations ( int max_iterations )

inline

Set the maximum number of iterations.

Parameters

[in] max_iterations Maximum number if iterations to run

Definition at line 175 of file correspondence_rejection_sample_consensus.h.

References max_iterations_.

Referenced by pcl::GeometricConsistencyGrouping< PointModelT, PointSceneT >::clusterCorrespondences(), and pcl::Hough3DGrouping< PointModelT, PointSceneT, PointModelRfT, PointSceneRfT >::clusterCorrespondences().

◆ setRefineModel()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setRefineModel ( const bool refine )

inline

Specify whether the model should be refined internally using the variance of the inliers.

Parameters

[in] refine true if the model should be refined, false otherwise

Definition at line 203 of file correspondence_rejection_sample_consensus.h.

References refine_.

◆ setSaveInliers()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setSaveInliers ( bool s )

inline

Set whether to save inliers or not.

Parameters

[in] s True to save inliers / False otherwise

Definition at line 230 of file correspondence_rejection_sample_consensus.h.

References save_inliers_.

◆ setSourcePoints()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setSourcePoints ( pcl::PCLPointCloud2::ConstPtr cloud2 )

inlineoverridevirtual

Blob method for setting the source cloud.

Reimplemented from pcl::registration::CorrespondenceRejector.

Definition at line 128 of file correspondence_rejection_sample_consensus.h.

References pcl::fromPCLPointCloud2(), and setInputSource().

◆ setTargetPoints()

template<typename PointT>

void pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::setTargetPoints ( pcl::PCLPointCloud2::ConstPtr cloud2 )

inlineoverridevirtual

template<typename PointT>

bool pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::save_inliers_ {false}

protected

Definition at line 264 of file correspondence_rejection_sample_consensus.h.

Referenced by getRemainingCorrespondences(), getSaveInliers(), and setSaveInliers().

◆ target_

template<typename PointT>

PointCloudConstPtr pcl::registration::CorrespondenceRejectorSampleConsensus< PointT >::target_

protected

Definition at line 258 of file correspondence_rejection_sample_consensus.h.

Referenced by CorrespondenceRejectorSampleConsensus(), getInputTarget(), pcl::registration::CorrespondenceRejectorSampleConsensus2D< PointT >::getRemainingCorrespondences(), getRemainingCorrespondences(), and setInputTarget().

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

pcl/registration/correspondence_rejection_sample_consensus.h
pcl/registration/impl/correspondence_rejection_sample_consensus.hpp

Public Types

Public Member Functions

Public Attributes

Protected Member Functions

Protected Attributes

Detailed Description

Member Typedef Documentation

◆ ConstPtr

◆ Ptr

Constructor & Destructor Documentation

◆ CorrespondenceRejectorSampleConsensus()

◆ ~CorrespondenceRejectorSampleConsensus()

Member Function Documentation

◆ applyRejection()

◆ getBestTransformation()

◆ getClassName()

◆ getInliersIndices()

◆ getInlierThreshold()

◆ getInputSource()

◆ getInputTarget()

◆ getMaximumIterations()

◆ getRefineModel()

◆ getRemainingCorrespondences()

◆ getSaveInliers()

◆ requiresSourcePoints()

◆ requiresTargetPoints()

◆ setInlierThreshold()

◆ setInputSource()

◆ setInputTarget()

◆ setMaximumIterations()

◆ setRefineModel()

◆ setSaveInliers()

◆ setSourcePoints()

◆ setTargetPoints()

Member Data Documentation

◆ best_transformation_

◆ inlier_indices_

◆ inlier_threshold_

◆ input_

◆ input_correspondences_

◆ input_transformed_

◆ max_iterations_

◆ refine_

◆ rejection_name_

◆ save_inliers_

◆ target_