pcl::SampleConsensusModelSphere< PointT > Class Template Reference

SampleConsensusModelSphere defines a model for 3D sphere segmentation. More...

#include <sac_model_sphere.h>

Inheritance diagram for pcl::SampleConsensusModelSphere< PointT >:
Inheritance graph
[legend]

List of all members.

Public Types

typedef SampleConsensusModel
< PointT >::PointCloud 
PointCloud
typedef SampleConsensusModel
< PointT >::PointCloudConstPtr 
PointCloudConstPtr
typedef SampleConsensusModel
< PointT >::PointCloudPtr 
PointCloudPtr
typedef boost::shared_ptr
< SampleConsensusModelSphere
Ptr

Public Member Functions

bool computeModelCoefficients (const std::vector< int > &samples, Eigen::VectorXf &model_coefficients)
 Check whether the given index samples can form a valid sphere model, compute the model coefficients from these samples and store them internally in model_coefficients. The sphere coefficients are: x, y, z, R.
bool doSamplesVerifyModel (const std::set< int > &indices, const Eigen::VectorXf &model_coefficients, double threshold)
 Verify whether a subset of indices verifies the given sphere model coefficients.
void getDistancesToModel (const Eigen::VectorXf &model_coefficients, std::vector< double > &distances)
 Compute all distances from the cloud data to a given sphere model.
pcl::SacModel getModelType () const
 Return an unique id for this model (SACMODEL_SPHERE).
void getSamples (int &iterations, std::vector< int > &samples)
 Get 4 random points (3 non-collinear) as data samples and return them as point indices.
void optimizeModelCoefficients (const std::vector< int > &inliers, const Eigen::VectorXf &model_coefficients, Eigen::VectorXf &optimized_coefficients)
 Recompute the sphere coefficients using the given inlier set and return them to the user.
void projectPoints (const std::vector< int > &inliers, const Eigen::VectorXf &model_coefficients, PointCloud &projected_points, bool copy_data_fields=true)
 Create a new point cloud with inliers projected onto the sphere model.
 SampleConsensusModelSphere (const PointCloudConstPtr &cloud, const std::vector< int > &indices)
 Constructor for base SampleConsensusModelSphere.
 SampleConsensusModelSphere (const PointCloudConstPtr &cloud)
 Constructor for base SampleConsensusModelSphere.
void selectWithinDistance (const Eigen::VectorXf &model_coefficients, double threshold, std::vector< int > &inliers)
 Select all the points which respect the given model coefficients as inliers.

Protected Member Functions

bool isModelValid (const Eigen::VectorXf &model_coefficients)
 Check whether a model is valid given the user constraints.

Static Private Member Functions

static int functionToOptimize (void *p, int m, int n, const double *x, double *fvec, int iflag)
 Cost function to be minimized.

Private Attributes

const std::vector< int > * tmp_inliers_
 Temporary pointer to a list of given indices for optimizeModelCoefficients ().
boost::mutex tmp_mutex_
 Temporary boost mutex for tmp_inliers_.

Static Private Attributes

static const int MAX_ITERATIONS_COLLINEAR = 1000
 Define the maximum number of iterations for collinearity checks.

Detailed Description

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

SampleConsensusModelSphere defines a model for 3D sphere segmentation.

Author:
Radu Bogdan Rusu

Definition at line 52 of file sac_model_sphere.h.


Member Typedef Documentation

template<typename PointT>
typedef SampleConsensusModel<PointT>::PointCloud pcl::SampleConsensusModelSphere< PointT >::PointCloud

Reimplemented from pcl::SampleConsensusModel< PointT >.

Definition at line 60 of file sac_model_sphere.h.

template<typename PointT>
typedef SampleConsensusModel<PointT>::PointCloudConstPtr pcl::SampleConsensusModelSphere< PointT >::PointCloudConstPtr

Reimplemented from pcl::SampleConsensusModel< PointT >.

Definition at line 62 of file sac_model_sphere.h.

template<typename PointT>
typedef SampleConsensusModel<PointT>::PointCloudPtr pcl::SampleConsensusModelSphere< PointT >::PointCloudPtr

Reimplemented from pcl::SampleConsensusModel< PointT >.

Definition at line 61 of file sac_model_sphere.h.

template<typename PointT>
typedef boost::shared_ptr<SampleConsensusModelSphere> pcl::SampleConsensusModelSphere< PointT >::Ptr

Reimplemented from pcl::SampleConsensusModel< PointT >.

Definition at line 64 of file sac_model_sphere.h.


Constructor & Destructor Documentation

template<typename PointT>
pcl::SampleConsensusModelSphere< PointT >::SampleConsensusModelSphere ( const PointCloudConstPtr cloud  )  [inline]

Constructor for base SampleConsensusModelSphere.

Parameters:
cloud the input point cloud dataset

Definition at line 69 of file sac_model_sphere.h.

template<typename PointT>
pcl::SampleConsensusModelSphere< PointT >::SampleConsensusModelSphere ( const PointCloudConstPtr cloud,
const std::vector< int > &  indices 
) [inline]

Constructor for base SampleConsensusModelSphere.

Parameters:
cloud the input point cloud dataset
indices a vector of point indices to be used from cloud

Definition at line 75 of file sac_model_sphere.h.


Member Function Documentation

template<typename PointT >
bool pcl::SampleConsensusModelSphere< PointT >::computeModelCoefficients ( const std::vector< int > &  samples,
Eigen::VectorXf &  model_coefficients 
) [inline, virtual]

Check whether the given index samples can form a valid sphere model, compute the model coefficients from these samples and store them internally in model_coefficients. The sphere coefficients are: x, y, z, R.

Parameters:
samples the point indices found as possible good candidates for creating a valid model
model_coefficients the resultant model coefficients

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 131 of file sac_model_sphere.hpp.

template<typename PointT >
bool pcl::SampleConsensusModelSphere< PointT >::doSamplesVerifyModel ( const std::set< int > &  indices,
const Eigen::VectorXf &  model_coefficients,
double  threshold 
) [inline, virtual]

Verify whether a subset of indices verifies the given sphere model coefficients.

Parameters:
indices the data indices that need to be tested against the sphere model
model_coefficients the sphere model coefficients
threshold a maximum admissible distance threshold for determining the inliers from the outliers

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 365 of file sac_model_sphere.hpp.

template<typename PointT >
int pcl::SampleConsensusModelSphere< PointT >::functionToOptimize ( void *  p,
int  m,
int  n,
const double *  x,
double *  fvec,
int  iflag 
) [inline, static, private]

Cost function to be minimized.

Parameters:
p a pointer to our data structure array
m the number of functions
n the number of variables
x a pointer to the variables array
fvec a pointer to the resultant functions evaluations
iflag set to -1 inside the function to terminate execution

Definition at line 321 of file sac_model_sphere.hpp.

template<typename PointT >
void pcl::SampleConsensusModelSphere< PointT >::getDistancesToModel ( const Eigen::VectorXf &  model_coefficients,
std::vector< double > &  distances 
) [inline, virtual]

Compute all distances from the cloud data to a given sphere model.

Parameters:
model_coefficients the coefficients of a sphere model that we need to compute distances to
distances the resultant estimated distances

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 198 of file sac_model_sphere.hpp.

template<typename PointT>
pcl::SacModel pcl::SampleConsensusModelSphere< PointT >::getModelType (  )  const [inline, virtual]

Return an unique id for this model (SACMODEL_SPHERE).

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 141 of file sac_model_sphere.h.

template<typename PointT >
void pcl::SampleConsensusModelSphere< PointT >::getSamples ( int &  iterations,
std::vector< int > &  samples 
) [inline, virtual]

Get 4 random points (3 non-collinear) as data samples and return them as point indices.

Parameters:
iterations the internal number of iterations used by SAC methods
samples the resultant model samples
Note:
assumes unique points!
Two different points could be enough in theory, to infere some sort of a center and a radius, but in practice, we might end up with a lot of points which are just 'close' to one another. Therefore we have two options: a) use normal information (good but I wouldn't rely on it in extremely noisy point clouds, no matter what) b) get two more points and uniquely identify a sphere in space (3 unique points define a circle)

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 45 of file sac_model_sphere.hpp.

template<typename PointT>
bool pcl::SampleConsensusModelSphere< PointT >::isModelValid ( const Eigen::VectorXf &  model_coefficients  )  [inline, protected, virtual]

Check whether a model is valid given the user constraints.

Parameters:
model_coefficients the set of model coefficients

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 148 of file sac_model_sphere.h.

template<typename PointT >
void pcl::SampleConsensusModelSphere< PointT >::optimizeModelCoefficients ( const std::vector< int > &  inliers,
const Eigen::VectorXf &  model_coefficients,
Eigen::VectorXf &  optimized_coefficients 
) [inline, virtual]

Recompute the sphere coefficients using the given inlier set and return them to the user.

Note:
: these are the coefficients of the sphere model after refinement (eg. after SVD)
Parameters:
inliers the data inliers found as supporting the model
model_coefficients the initial guess for the optimization
optimized_coefficients the resultant recomputed coefficients after non-linear optimization

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 266 of file sac_model_sphere.hpp.

template<typename PointT >
void pcl::SampleConsensusModelSphere< PointT >::projectPoints ( const std::vector< int > &  inliers,
const Eigen::VectorXf &  model_coefficients,
PointCloud projected_points,
bool  copy_data_fields = true 
) [inline, virtual]

Create a new point cloud with inliers projected onto the sphere model.

Parameters:
inliers the data inliers that we want to project on the sphere model
model_coefficients the coefficients of a sphere model
projected_points the resultant projected points
copy_data_fields set to true if we need to copy the other data fields
Todo:
implement this.

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 342 of file sac_model_sphere.hpp.

template<typename PointT >
void pcl::SampleConsensusModelSphere< PointT >::selectWithinDistance ( const Eigen::VectorXf &  model_coefficients,
double  threshold,
std::vector< int > &  inliers 
) [inline, virtual]

Select all the points which respect the given model coefficients as inliers.

Parameters:
model_coefficients the coefficients of a sphere model that we need to compute distances to
threshold a maximum admissible distance threshold for determining the inliers from the outliers
inliers the resultant model inliers

Implements pcl::SampleConsensusModel< PointT >.

Definition at line 227 of file sac_model_sphere.hpp.


Member Data Documentation

template<typename PointT>
const int pcl::SampleConsensusModelSphere< PointT >::MAX_ITERATIONS_COLLINEAR = 1000 [static, private]

Define the maximum number of iterations for collinearity checks.

Definition at line 173 of file sac_model_sphere.h.

template<typename PointT>
const std::vector<int>* pcl::SampleConsensusModelSphere< PointT >::tmp_inliers_ [private]

Temporary pointer to a list of given indices for optimizeModelCoefficients ().

Definition at line 170 of file sac_model_sphere.h.

template<typename PointT>
boost::mutex pcl::SampleConsensusModelSphere< PointT >::tmp_mutex_ [private]

Temporary boost mutex for tmp_inliers_.

Definition at line 167 of file sac_model_sphere.h.


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines


pcl
Author(s): See http://pcl.ros.org/authors for the complete list of authors.
autogenerated on Fri Jan 11 09:57:23 2013