Principal Component analysis (PCA) class. More...

#include <pcl/common/pca.h>

Inheritance diagram for pcl::PCA< PointT >:

Collaboration diagram for pcl::PCA< PointT >:

Public Types
enum	FLAG { increase , preserve }
	Updating method flag. More...

using	Base = pcl::PCLBase< PointT >

using	PointCloud = typename Base::PointCloud

using	PointCloudPtr = typename Base::PointCloudPtr

using	PointCloudConstPtr = typename Base::PointCloudConstPtr

using	PointIndicesPtr = typename Base::PointIndicesPtr

using	PointIndicesConstPtr = typename Base::PointIndicesConstPtr

Public Types inherited from pcl::PCLBase< PointT >
using	PointCloud = pcl::PointCloud< PointT >

using	PointCloudPtr = typename PointCloud::Ptr

using	PointCloudConstPtr = typename PointCloud::ConstPtr

using	PointIndicesPtr = PointIndices::Ptr

using	PointIndicesConstPtr = PointIndices::ConstPtr

Public Member Functions
	PCA (bool basis_only=false)
	Default Constructor. More...

	PCA (PCA const &pca)
	Copy Constructor. More...

PCA &	operator= (PCA const &pca)
	Assignment operator. More...

void	setInputCloud (const PointCloudConstPtr &cloud) override
	Provide a pointer to the input dataset. More...

void	setIndices (const IndicesPtr &indices) override
	Provide a pointer to the vector of indices that represents the input data. More...

void	setIndices (const IndicesConstPtr &indices) override
	Provide a pointer to the vector of indices that represents the input data. More...

void	setIndices (const PointIndicesConstPtr &indices) override
	Provide a pointer to the vector of indices that represents the input data. More...

void	setIndices (std::size_t row_start, std::size_t col_start, std::size_t nb_rows, std::size_t nb_cols) override
	Set the indices for the points laying within an interest region of the point cloud. More...

Eigen::Vector4f &	getMean ()
	Mean accessor. More...

Eigen::Matrix3f &	getEigenVectors ()
	Eigen Vectors accessor. More...

Eigen::Vector3f &	getEigenValues ()
	Eigen Values accessor. More...

Eigen::MatrixXf &	getCoefficients ()
	Coefficients accessor. More...

void	update (const PointT &input, FLAG flag=preserve)
	update PCA with a new point More...

void	project (const PointT &input, PointT &projection)
	Project point on the eigenspace. More...

void	project (const PointCloud &input, PointCloud &projection)
	Project cloud on the eigenspace. More...

void	reconstruct (const PointT &projection, PointT &input)
	Reconstruct point from its projection. More...

void	reconstruct (const PointCloud &projection, PointCloud &input)
	Reconstruct cloud from its projection. More...

bool	initCompute ()
	This method should get called before starting the actual computation. More...

virtual void	setInputCloud (const PointCloudConstPtr &cloud)
	Provide a pointer to the input dataset. More...

Public Member Functions inherited from pcl::PCLBase< PointT >
	PCLBase ()
	Empty constructor. More...

	PCLBase (const PCLBase &base)
	Copy constructor. More...

virtual	~PCLBase ()=default
	Destructor. More...

PointCloudConstPtr const	getInputCloud () const
	Get a pointer to the input point cloud dataset. More...

IndicesPtr	getIndices ()
	Get a pointer to the vector of indices used. More...

IndicesConstPtr const	getIndices () const
	Get a pointer to the vector of indices used. More...

const PointT &	operator[] (std::size_t pos) const
	Override PointCloud operator[] to shorten code. More...

Public Attributes
PointCloudConstPtr	input_
	The input point cloud dataset. More...

IndicesPtr	indices_
	A pointer to the vector of point indices to use. More...

Additional Inherited Members
Protected Member Functions inherited from pcl::PCLBase< PointT >
bool	initCompute ()
	This method should get called before starting the actual computation. More...

bool	deinitCompute ()
	This method should get called after finishing the actual computation. More...

Protected Attributes inherited from pcl::PCLBase< PointT >
PointCloudConstPtr	input_
	The input point cloud dataset. More...

IndicesPtr	indices_
	A pointer to the vector of point indices to use. More...

bool	use_indices_
	Set to true if point indices are used. More...

bool	fake_indices_
	If no set of indices are given, we construct a set of fake indices that mimic the input PointCloud. More...

Detailed Description

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

Principal Component analysis (PCA) class.

Principal components are extracted by singular values decomposition on the covariance matrix of the centered input cloud. Available data after pca computation are:

The Mean of the input data
The Eigenvectors: Ordered set of vectors representing the resultant principal components and the eigenspace cartesian basis (right-handed coordinate system).
The Eigenvalues: Eigenvectors correspondent loadings ordered in descending order.

Other methods allow projection in the eigenspace, reconstruction from eigenspace and update of the eigenspace with a new datum (according Matej Artec, Matjaz Jogan and Ales Leonardis: "Incremental PCA for On-line Visual Learning and Recognition").

Author: Nizar Sallem

Definition at line 61 of file pca.h.

Member Typedef Documentation

◆ Base

template<typename PointT >

using pcl::PCA< PointT >::Base = pcl::PCLBase<PointT>

Definition at line 64 of file pca.h.

◆ PointCloud

template<typename PointT >

using pcl::PCA< PointT >::PointCloud = typename Base::PointCloud

Definition at line 65 of file pca.h.

◆ PointCloudConstPtr

template<typename PointT >

using pcl::PCA< PointT >::PointCloudConstPtr = typename Base::PointCloudConstPtr

Definition at line 67 of file pca.h.

◆ PointCloudPtr

template<typename PointT >

using pcl::PCA< PointT >::PointCloudPtr = typename Base::PointCloudPtr

Definition at line 66 of file pca.h.

◆ PointIndicesConstPtr

template<typename PointT >

using pcl::PCA< PointT >::PointIndicesConstPtr = typename Base::PointIndicesConstPtr

Definition at line 69 of file pca.h.

◆ PointIndicesPtr

template<typename PointT >

using pcl::PCA< PointT >::PointIndicesPtr = typename Base::PointIndicesPtr

Definition at line 68 of file pca.h.

Member Enumeration Documentation

◆ FLAG

template<typename PointT >

enum pcl::PCA::FLAG

Updating method flag.

Enumerator
increase	keep the new basis vectors if possible
preserve	preserve subspace dimension

Definition at line 77 of file pca.h.

Constructor & Destructor Documentation

◆ PCA() [1/2]

template<typename PointT >

pcl::PCA< PointT >::PCA ( bool basis_only = false )

inline

Default Constructor.

Parameters

basis_only flag to compute only the PCA basis

Definition at line 88 of file pca.h.

◆ PCA() [2/2]

template<typename PointT >

pcl::PCA< PointT >::PCA ( PCA< PointT > const & pca )

inline

Copy Constructor.

Parameters

[in] pca PCA object

Definition at line 96 of file pca.h.

Member Function Documentation

◆ getCoefficients()

template<typename PointT >

Eigen::MatrixXf& pcl::PCA< PointT >::getCoefficients ( )

inline

Coefficients accessor.

Exceptions

InitFailedException

Definition at line 221 of file pca.h.

◆ getEigenValues()

template<typename PointT >

Eigen::Vector3f& pcl::PCA< PointT >::getEigenValues ( )

inline

Eigen Values accessor.

Exceptions

InitFailedException

Definition at line 207 of file pca.h.

◆ getEigenVectors()

template<typename PointT >

Eigen::Matrix3f& pcl::PCA< PointT >::getEigenVectors ( )

inline

Eigen Vectors accessor.

Returns: Column ordered eigenvectors, representing the eigenspace cartesian basis (right-handed coordinate system).

Exceptions

InitFailedException

Definition at line 193 of file pca.h.

◆ getMean()

template<typename PointT >

Eigen::Vector4f& pcl::PCA< PointT >::getMean ( )

inline

Mean accessor.

Exceptions

InitFailedException

Definition at line 178 of file pca.h.

◆ initCompute()

template<typename PointT >

bool pcl::PCLBase< PointT >::initCompute

This method should get called before starting the actual computation.

Internally, initCompute() does the following:

checks if an input dataset is given, and returns false otherwise
checks whether a set of input indices has been given. Returns true if yes.
if no input indices have been given, a fake set is created, which will be used until:
- either a new set is given via setIndices(), or
- a new cloud is given that has a different set of points. This will trigger an update on the set of fake indices

Definition at line 168 of file pcl_base.hpp.

◆ operator=()

template<typename PointT >

PCA& pcl::PCA< PointT >::operator= ( PCA< PointT > const & pca )

inline

Assignment operator.

Parameters

[in] pca PCA object

Definition at line 110 of file pca.h.

◆ project() [1/2]

template<typename PointT >

void pcl::PCA< PointT >::project	(	const PointCloud &	input,
		PointCloud &	projection
	)

inline

Project cloud on the eigenspace.

Parameters

[in]	input	cloud from original dataset
[out]	projection	the cloud in eigen vectors space

Exceptions

InitFailedException

Definition at line 179 of file pca.hpp.

References pcl::geometry::project().

◆ project() [2/2]

template<typename PointT >

void pcl::PCA< PointT >::project	(	const PointT &	input,
		PointT &	projection
	)

inline

Project point on the eigenspace.

Parameters

[in]	input	point from original dataset
[out]	projection	the point in eigen vectors space

Exceptions

InitFailedException

Definition at line 166 of file pca.hpp.

◆ reconstruct() [1/2]

template<typename PointT >

void pcl::PCA< PointT >::reconstruct	(	const PointCloud &	projection,
		PointCloud &	input
	)

inline

Reconstruct cloud from its projection.

Parameters

[in]	projection	cloud from eigenvector space
[out]	input	reconstructed cloud

Exceptions

InitFailedException

Definition at line 221 of file pca.hpp.

◆ reconstruct() [2/2]

template<typename PointT >

void pcl::PCA< PointT >::reconstruct	(	const PointT &	projection,
		PointT &	input
	)

inline

Reconstruct point from its projection.

Parameters

[in]	projection	point from eigenvector space
[out]	input	reconstructed point

Exceptions

InitFailedException

Definition at line 208 of file pca.hpp.

◆ setIndices() [1/4]

template<typename PointT >

void pcl::PCA< PointT >::setIndices ( const IndicesConstPtr & indices )

inlineoverridevirtual

Provide a pointer to the vector of indices that represents the input data.

Parameters

[in] indices a pointer to the indices that represent the input data.

Reimplemented from pcl::PCLBase< PointT >.

Definition at line 143 of file pca.h.

References pcl::PCLBase< PointT >::setIndices().

◆ setIndices() [2/4]

template<typename PointT >

void pcl::PCA< PointT >::setIndices ( const IndicesPtr & indices )

inlineoverridevirtual

Provide a pointer to the vector of indices that represents the input data.

Parameters

[in] indices a pointer to the indices that represent the input data.

Reimplemented from pcl::PCLBase< PointT >.

Definition at line 133 of file pca.h.

References pcl::PCLBase< PointT >::setIndices().

◆ setIndices() [3/4]

template<typename PointT >

void pcl::PCA< PointT >::setIndices ( const PointIndicesConstPtr & indices )

inlineoverridevirtual

Provide a pointer to the vector of indices that represents the input data.

Parameters

[in] indices a pointer to the indices that represent the input data.

Reimplemented from pcl::PCLBase< PointT >.

Definition at line 153 of file pca.h.

References pcl::PCLBase< PointT >::setIndices().

◆ setIndices() [4/4]

template<typename PointT >

void pcl::PCA< PointT >::setIndices	(	std::size_t	row_start,
		std::size_t	col_start,
		std::size_t	nb_rows,
		std::size_t	nb_cols
	)

inlineoverridevirtual

Set the indices for the points laying within an interest region of the point cloud.

Note: you shouldn't call this method on unorganized point clouds!

Parameters

[in]	row_start	the offset on rows
[in]	col_start	the offset on columns
[in]	nb_rows	the number of rows to be considered row_start included
[in]	nb_cols	the number of columns to be considered col_start included

Reimplemented from pcl::PCLBase< PointT >.

Definition at line 168 of file pca.h.

References pcl::PCLBase< PointT >::setIndices().

◆ setInputCloud() [1/2]

template<typename PointT >

void pcl::PCLBase< PointT >::setInputCloud

Provide a pointer to the input dataset.

Parameters

[in] cloud the const boost shared pointer to a PointCloud message

Definition at line 92 of file pcl_base.hpp.

◆ setInputCloud() [2/2]

template<typename PointT >

void pcl::PCA< PointT >::setInputCloud ( const PointCloudConstPtr & cloud )

inlineoverridevirtual

Provide a pointer to the input dataset.

Parameters

cloud the const boost shared pointer to a PointCloud message

Reimplemented from pcl::PCLBase< PointT >.

Definition at line 123 of file pca.h.

References pcl::PCLBase< PointT >::setInputCloud().

◆ update()

template<typename PointT >

void pcl::PCA< PointT >::update	(	const PointT &	input,
		FLAG	flag = `preserve`
	)

inline

update PCA with a new point

Parameters

[in]	input	input point
[in]	flag	update flag

Exceptions

InitFailedException

Definition at line 96 of file pca.hpp.

Member Data Documentation

◆ indices_

template<typename PointT >

IndicesPtr pcl::PCLBase< PointT >::indices_

A pointer to the vector of point indices to use.

Definition at line 150 of file pcl_base.h.

◆ input_

template<typename PointT >

PointCloudConstPtr pcl::PCLBase< PointT >::input_

The input point cloud dataset.

Definition at line 147 of file pcl_base.h.

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

pcl/common/pca.h
pcl/common/impl/pca.hpp

Public Types

Public Member Functions

Public Attributes

Additional Inherited Members

Detailed Description

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

Member Typedef Documentation

◆ Base

◆ PointCloud

◆ PointCloudConstPtr

◆ PointCloudPtr

◆ PointIndicesConstPtr

◆ PointIndicesPtr

Member Enumeration Documentation

◆ FLAG

Constructor & Destructor Documentation

◆ PCA() [1/2]

◆ PCA() [2/2]

Member Function Documentation

◆ getCoefficients()

◆ getEigenValues()

◆ getEigenVectors()

◆ getMean()

◆ initCompute()

◆ operator=()

◆ project() [1/2]

◆ project() [2/2]

◆ reconstruct() [1/2]

◆ reconstruct() [2/2]

◆ setIndices() [1/4]

◆ setIndices() [2/4]

◆ setIndices() [3/4]

◆ setIndices() [4/4]

◆ setInputCloud() [1/2]

◆ setInputCloud() [2/2]

◆ update()

Member Data Documentation

◆ indices_

◆ input_

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