Implements a supervoxel algorithm based on voxel structure, normals, and rgb values. More...

#include <supervoxel_clustering.h>

Inheritance diagram for pcl::SupervoxelClustering< PointT >:

Classes
class	SupervoxelHelper
	Internal storage class for supervoxels. More...
class	VoxelData
	VoxelData is a structure used for storing data within a pcl::octree::OctreePointCloudAdjacencyContainer. More...
Public Types
typedef VoxelAdjacencyList::edge_descriptor	EdgeID
typedef boost::shared_ptr < std::vector< int > >	IndicesPtr
typedef pcl::search::KdTree < PointT >	KdTreeT
typedef pcl::octree::OctreePointCloudAdjacencyContainer < PointT, VoxelData >	LeafContainerT
typedef std::vector < LeafContainerT * >	LeafVectorT
typedef pcl::PointCloud< Normal >	NormalCloudT
typedef pcl::octree::OctreePointCloudAdjacency < PointT, LeafContainerT >	OctreeAdjacencyT
typedef pcl::octree::OctreePointCloudSearch < PointT >	OctreeSearchT
typedef pcl::PointCloud< PointT >	PointCloudT
typedef boost::adjacency_list < boost::setS, boost::setS, boost::undirectedS, uint32_t, float >	VoxelAdjacencyList
typedef VoxelAdjacencyList::vertex_descriptor	VoxelID
Public Member Functions
virtual void	extract (std::map< uint32_t, typename Supervoxel< PointT >::Ptr > &supervoxel_clusters)
	This method launches the segmentation algorithm and returns the supervoxels that were obtained during the segmentation.
pcl::PointCloud< PointXYZRGBA > ::Ptr	getColoredCloud () const
	Returns an RGB colorized cloud showing superpixels Otherwise it returns an empty pointer. Points that belong to the same supervoxel have the same color. But this function doesn't guarantee that different segments will have different color(it's random). Points that are unlabeled will be black.
pcl::PointCloud < pcl::PointXYZRGBA >::Ptr	getColoredVoxelCloud () const
	Returns an RGB colorized voxelized cloud showing superpixels Otherwise it returns an empty pointer. Points that belong to the same supervoxel have the same color. But this function doesn't guarantee that different segments will have different color(it's random). Points that are unlabeled will be black.
pcl::PointCloud< PointXYZL >::Ptr	getLabeledCloud () const
	Returns labeled cloud Points that belong to the same supervoxel have the same label. Labels for segments start from 1, unlabled points have label 0.
pcl::PointCloud < pcl::PointXYZL >::Ptr	getLabeledVoxelCloud () const
	Returns labeled voxelized cloud Points that belong to the same supervoxel have the same label. Labels for segments start from 1, unlabled points have label 0.
float	getSeedResolution () const
	Get the resolution of the octree seed voxels.
void	getSupervoxelAdjacency (std::multimap< uint32_t, uint32_t > &label_adjacency) const
	Get a multimap which gives supervoxel adjacency.
void	getSupervoxelAdjacencyList (VoxelAdjacencyList &adjacency_list_arg) const
	Gets the adjacency list (Boost Graph library) which gives connections between supervoxels.
pcl::PointCloud< PointT >::Ptr	getVoxelCentroidCloud () const
	Returns a deep copy of the voxel centroid cloud.
float	getVoxelResolution () const
	Get the resolution of the octree voxels.
virtual void	refineSupervoxels (int num_itr, std::map< uint32_t, typename Supervoxel< PointT >::Ptr > &supervoxel_clusters)
	This method refines the calculated supervoxels - may only be called after extract.
void	setColorImportance (float val)
	Set the importance of color for supervoxels.
virtual void	setInputCloud (typename pcl::PointCloud< PointT >::ConstPtr cloud)
	This method sets the cloud to be supervoxelized.
void	setNormalImportance (float val)
	Set the importance of scalar normal product for supervoxels.
void	setSeedResolution (float seed_resolution)
	Set the resolution of the octree seed voxels.
void	setSpatialImportance (float val)
	Set the importance of spatial distance for supervoxels.
void	setVoxelResolution (float resolution)
	Set the resolution of the octree voxels.
	SupervoxelClustering (float voxel_resolution, float seed_resolution, bool use_single_camera_transform=true)
	Constructor that sets default values for member variables.
virtual	~SupervoxelClustering ()
	This destructor destroys the cloud, normals and search method used for finding neighbors. In other words it frees memory.
Static Public Member Functions
static pcl::PointCloud < pcl::PointNormal >::Ptr	makeSupervoxelNormalCloud (std::map< uint32_t, typename Supervoxel< PointT >::Ptr > &supervoxel_clusters)
	Static helper function which returns a pointcloud of normals for the input supervoxels.
Private Types
typedef boost::ptr_list < SupervoxelHelper >	HelperListT
Private Member Functions
void	computeVoxelData ()
	This sets the data of the voxels in the tree.
void	createSupervoxelHelpers (std::vector< PointT, Eigen::aligned_allocator< PointT > > &seed_points)
	This method creates the internal supervoxel helpers based on the provided seed points.
void	expandSupervoxels (int depth)
	This performs the superpixel evolution.
void	makeSupervoxels (std::map< uint32_t, typename Supervoxel< PointT >::Ptr > &supervoxel_clusters)
	Constructs the map of supervoxel clusters from the internal supervoxel helpers.
virtual bool	prepareForSegmentation ()
	This method simply checks if it is possible to execute the segmentation algorithm with the current settings. If it is possible then it returns true.
void	reseedSupervoxels ()
	Reseeds the supervoxels by finding the voxel closest to current centroid.
void	selectInitialSupervoxelSeeds (std::vector< PointT, Eigen::aligned_allocator< PointT > > &seed_points)
	This selects points to use as initial supervoxel centroids.
void	transformFunction (PointT &p)
	Transform function used to normalize voxel density versus distance from camera.
float	voxelDataDistance (const VoxelData &v1, const VoxelData &v2) const
	Distance function used for comparing voxelDatas.
Private Attributes
OctreeAdjacencyT::Ptr	adjacency_octree_
	Octree Adjacency structure with leaves at voxel resolution.
float	color_importance_
	Importance of color in clustering.
std::vector< uint32_t >	label_colors_
	Stores the colors used for the superpixel labels.
float	normal_importance_
	Importance of similarity in normals for clustering.
float	resolution_
	Stores the resolution used in the octree.
float	seed_resolution_
	Stores the resolution used to seed the superpixels.
float	spatial_importance_
	Importance of distance from seed center in clustering.
HelperListT	supervoxel_helpers_
StopWatch	timer_
PointCloudT::Ptr	voxel_centroid_cloud_
	Contains the Voxelized centroid Cloud.
pcl::search::KdTree< PointT >::Ptr	voxel_kdtree_
	Contains a KDtree for the voxelized cloud.
Friends
class	SupervoxelHelper

Detailed Description

template<typename PointT>
class pcl::SupervoxelClustering< PointT >

Implements a supervoxel algorithm based on voxel structure, normals, and rgb values.

Note:

Supervoxels are oversegmented volumetric patches (usually surfaces)

Usually, color isn't needed (and can be detrimental)- spatial structure is mainly used

J. Papon, A. Abramov, M. Schoeler, F. Woergoetter Voxel Cloud Connectivity Segmentation - Supervoxels from PointClouds In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2013

Author:: Jeremie Papon (jpapon@gmail.com)

Definition at line 119 of file supervoxel_clustering.h.

Member Typedef Documentation

template<typename PointT>

typedef VoxelAdjacencyList::edge_descriptor pcl::SupervoxelClustering< PointT >::EdgeID

Definition at line 177 of file supervoxel_clustering.h.

template<typename PointT>

typedef boost::ptr_list<SupervoxelHelper> pcl::SupervoxelClustering< PointT >::HelperListT [private]

Definition at line 463 of file supervoxel_clustering.h.

template<typename PointT>

typedef boost::shared_ptr<std::vector<int> > pcl::SupervoxelClustering< PointT >::IndicesPtr

Definition at line 169 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::search::KdTree<PointT> pcl::SupervoxelClustering< PointT >::KdTreeT

Definition at line 168 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::octree::OctreePointCloudAdjacencyContainer<PointT, VoxelData> pcl::SupervoxelClustering< PointT >::LeafContainerT

Definition at line 161 of file supervoxel_clustering.h.

template<typename PointT>

typedef std::vector<LeafContainerT*> pcl::SupervoxelClustering< PointT >::LeafVectorT

Definition at line 162 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::PointCloud<Normal> pcl::SupervoxelClustering< PointT >::NormalCloudT

Definition at line 165 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::octree::OctreePointCloudAdjacency<PointT, LeafContainerT> pcl::SupervoxelClustering< PointT >::OctreeAdjacencyT

Definition at line 166 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::octree::OctreePointCloudSearch<PointT> pcl::SupervoxelClustering< PointT >::OctreeSearchT

Definition at line 167 of file supervoxel_clustering.h.

template<typename PointT>

typedef pcl::PointCloud<PointT> pcl::SupervoxelClustering< PointT >::PointCloudT

Definition at line 164 of file supervoxel_clustering.h.

template<typename PointT>

typedef boost::adjacency_list<boost::setS, boost::setS, boost::undirectedS, uint32_t, float> pcl::SupervoxelClustering< PointT >::VoxelAdjacencyList

Definition at line 175 of file supervoxel_clustering.h.

template<typename PointT>

typedef VoxelAdjacencyList::vertex_descriptor pcl::SupervoxelClustering< PointT >::VoxelID

Definition at line 176 of file supervoxel_clustering.h.

Constructor & Destructor Documentation

template<typename PointT >

pcl::SupervoxelClustering< PointT >::SupervoxelClustering	(	float	voxel_resolution,
		float	seed_resolution,
		bool	use_single_camera_transform = `true`
	)

Constructor that sets default values for member variables.

Parameters:

[in]	voxel_resolution	The resolution (in meters) of voxels used
[in]	seed_resolution	The average size (in meters) of resulting supervoxels
[in]	use_single_camera_transform	Set to true if point density in cloud falls off with distance from origin (such as with a cloud coming from one stationary camera), set false if input cloud is from multiple captures from multiple locations.

Definition at line 47 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::SupervoxelClustering< PointT >::~SupervoxelClustering ( ) [virtual]

This destructor destroys the cloud, normals and search method used for finding neighbors. In other words it frees memory.

Definition at line 76 of file supervoxel_clustering.hpp.

Member Function Documentation

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::computeVoxelData ( ) [private]

This sets the data of the voxels in the tree.

Definition at line 207 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::createSupervoxelHelpers ( std::vector< PointT, Eigen::aligned_allocator< PointT > > & seed_points ) [private]

This method creates the internal supervoxel helpers based on the provided seed points.

Parameters:

[in] seed_points The selected points

Definition at line 308 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::expandSupervoxels ( int depth ) [private]

This performs the superpixel evolution.

Definition at line 258 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::extract ( std::map< uint32_t, typename Supervoxel< PointT >::Ptr > & supervoxel_clusters ) [virtual]

This method launches the segmentation algorithm and returns the supervoxels that were obtained during the segmentation.

Parameters:

[out] supervoxel_clusters A map of labels to pointers to supervoxel structures

Definition at line 98 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< pcl::PointXYZRGBA >::Ptr pcl::SupervoxelClustering< PointT >::getColoredCloud ( ) const

Returns an RGB colorized cloud showing superpixels Otherwise it returns an empty pointer. Points that belong to the same supervoxel have the same color. But this function doesn't guarantee that different segments will have different color(it's random). Points that are unlabeled will be black.

Definition at line 514 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< pcl::PointXYZRGBA >::Ptr pcl::SupervoxelClustering< PointT >::getColoredVoxelCloud ( ) const

Returns an RGB colorized voxelized cloud showing superpixels Otherwise it returns an empty pointer. Points that belong to the same supervoxel have the same color. But this function doesn't guarantee that different segments will have different color(it's random). Points that are unlabeled will be black.

Definition at line 545 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< pcl::PointXYZL >::Ptr pcl::SupervoxelClustering< PointT >::getLabeledCloud ( ) const

Returns labeled cloud Points that belong to the same supervoxel have the same label. Labels for segments start from 1, unlabled points have label 0.

Definition at line 598 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< pcl::PointXYZL >::Ptr pcl::SupervoxelClustering< PointT >::getLabeledVoxelCloud ( ) const

Returns labeled voxelized cloud Points that belong to the same supervoxel have the same label. Labels for segments start from 1, unlabled points have label 0.

Definition at line 576 of file supervoxel_clustering.hpp.

template<typename PointT >

float pcl::SupervoxelClustering< PointT >::getSeedResolution ( ) const

Get the resolution of the octree seed voxels.

Definition at line 660 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::getSupervoxelAdjacency ( std::multimap< uint32_t, uint32_t > & label_adjacency ) const

Get a multimap which gives supervoxel adjacency.

Parameters:

[out] label_adjacency Multi-Map which maps a supervoxel label to all adjacent supervoxel labels

Definition at line 495 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::getSupervoxelAdjacencyList ( VoxelAdjacencyList & adjacency_list_arg ) const

Gets the adjacency list (Boost Graph library) which gives connections between supervoxels.

Parameters:

[out] adjacency_list_arg BGL graph where supervoxel labels are vertices, edges are touching relationships

Definition at line 444 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< PointT >::Ptr pcl::SupervoxelClustering< PointT >::getVoxelCentroidCloud ( ) const

Returns a deep copy of the voxel centroid cloud.

Definition at line 567 of file supervoxel_clustering.hpp.

template<typename PointT >

float pcl::SupervoxelClustering< PointT >::getVoxelResolution ( ) const

Get the resolution of the octree voxels.

Definition at line 645 of file supervoxel_clustering.hpp.

template<typename PointT >

pcl::PointCloud< pcl::PointNormal >::Ptr pcl::SupervoxelClustering< PointT >::makeSupervoxelNormalCloud ( std::map< uint32_t, typename Supervoxel< PointT >::Ptr > & supervoxel_clusters ) [static]

Static helper function which returns a pointcloud of normals for the input supervoxels.

Parameters:

[in] supervoxel_clusters Supervoxel cluster map coming from this class

Returns:: Cloud of PointNormals of the supervoxels

Definition at line 628 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::makeSupervoxels ( std::map< uint32_t, typename Supervoxel< PointT >::Ptr > & supervoxel_clusters ) [private]

Constructs the map of supervoxel clusters from the internal supervoxel helpers.

Definition at line 290 of file supervoxel_clustering.hpp.

template<typename PointT >

bool pcl::SupervoxelClustering< PointT >::prepareForSegmentation ( ) [private, virtual]

This method simply checks if it is possible to execute the segmentation algorithm with the current settings. If it is possible then it returns true.

Definition at line 182 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::refineSupervoxels	(	int	num_itr,
		std::map< uint32_t, typename Supervoxel< PointT >::Ptr > &	supervoxel_clusters
	)		`[virtual]`

This method refines the calculated supervoxels - may only be called after extract.

Parameters:

[in]	num_itr	The number of iterations of refinement to be done (2 or 3 is usually sufficient)
[out]	supervoxel_clusters	The resulting refined supervoxels

Definition at line 150 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::reseedSupervoxels ( ) [private]

Reseeds the supervoxels by finding the voxel closest to current centroid.

Definition at line 388 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::selectInitialSupervoxelSeeds ( std::vector< PointT, Eigen::aligned_allocator< PointT > > & seed_points ) [private]

This selects points to use as initial supervoxel centroids.

Parameters:

[out] seed_points The selected points

Definition at line 330 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setColorImportance ( float val )

Set the importance of color for supervoxels.

Definition at line 675 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setInputCloud ( typename pcl::PointCloud< PointT >::ConstPtr cloud ) [virtual]

This method sets the cloud to be supervoxelized.

Parameters:

[in] cloud The cloud to be supervoxelize

Definition at line 83 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setNormalImportance ( float val )

Set the importance of scalar normal product for supervoxels.

Definition at line 689 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setSeedResolution ( float seed_resolution )

Set the resolution of the octree seed voxels.

Definition at line 667 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setSpatialImportance ( float val )

Set the importance of spatial distance for supervoxels.

Definition at line 682 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::setVoxelResolution ( float resolution )

Set the resolution of the octree voxels.

Definition at line 652 of file supervoxel_clustering.hpp.

template<typename PointT >

void pcl::SupervoxelClustering< PointT >::transformFunction ( PointT & p ) [private]

Transform function used to normalize voxel density versus distance from camera.

Definition at line 416 of file supervoxel_clustering.hpp.

template<typename PointT >

float pcl::SupervoxelClustering< PointT >::voxelDataDistance	(	const VoxelData &	v1,
		const VoxelData &	v2
	)		const `[private]`

Distance function used for comparing voxelDatas.

Definition at line 425 of file supervoxel_clustering.hpp.

Friends And Related Function Documentation

template<typename PointT>

friend class SupervoxelHelper [friend]

Definition at line 122 of file supervoxel_clustering.h.

Member Data Documentation

template<typename PointT>

OctreeAdjacencyT::Ptr pcl::SupervoxelClustering< PointT >::adjacency_octree_ [private]

Octree Adjacency structure with leaves at voxel resolution.

Definition at line 353 of file supervoxel_clustering.h.

template<typename PointT>

float pcl::SupervoxelClustering< PointT >::color_importance_ [private]

Importance of color in clustering.

Definition at line 359 of file supervoxel_clustering.h.

template<typename PointT>

std::vector<uint32_t> pcl::SupervoxelClustering< PointT >::label_colors_ [private]

Stores the colors used for the superpixel labels.

Definition at line 366 of file supervoxel_clustering.h.

template<typename PointT>

float pcl::SupervoxelClustering< PointT >::normal_importance_ [private]

Importance of similarity in normals for clustering.

Definition at line 363 of file supervoxel_clustering.h.

template<typename PointT>

float pcl::SupervoxelClustering< PointT >::resolution_ [private]

Stores the resolution used in the octree.

Definition at line 336 of file supervoxel_clustering.h.

template<typename PointT>

float pcl::SupervoxelClustering< PointT >::seed_resolution_ [private]

Stores the resolution used to seed the superpixels.

Definition at line 339 of file supervoxel_clustering.h.

template<typename PointT>

float pcl::SupervoxelClustering< PointT >::spatial_importance_ [private]

Importance of distance from seed center in clustering.

Definition at line 361 of file supervoxel_clustering.h.

template<typename PointT>

HelperListT pcl::SupervoxelClustering< PointT >::supervoxel_helpers_ [private]

Definition at line 464 of file supervoxel_clustering.h.

template<typename PointT>

StopWatch pcl::SupervoxelClustering< PointT >::timer_ [private]

Definition at line 467 of file supervoxel_clustering.h.

template<typename PointT>

PointCloudT::Ptr pcl::SupervoxelClustering< PointT >::voxel_centroid_cloud_ [private]

Contains the Voxelized centroid Cloud.

Definition at line 356 of file supervoxel_clustering.h.

template<typename PointT>

pcl::search::KdTree<PointT>::Ptr pcl::SupervoxelClustering< PointT >::voxel_kdtree_ [private]

Contains a KDtree for the voxelized cloud.

Definition at line 350 of file supervoxel_clustering.h.

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

Classes

Public Types

Public Member Functions

Static Public Member Functions

Private Types

Private Member Functions

Private Attributes

Friends

Detailed Description

template<typename PointT> class pcl::SupervoxelClustering< PointT >

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation

Friends And Related Function Documentation

Member Data Documentation

template<typename PointT>
class pcl::SupervoxelClustering< PointT >