Normal-based feature signature estimation class. Obtains the feature vector by applying Discrete Cosine and Fourier Transforms on an NxM array of real numbers representing the projection distances of the points in the input cloud to a disc around the point of interest. Please consult the following publication for more details: Xinju Li and Igor Guskov Multi-scale features for approximate alignment of point-based surfaces Proceedings of the third Eurographics symposium on Geometry processing July 2005, Vienna, Austria. More...
#include <normal_based_signature.h>
Public Types | |
typedef boost::shared_ptr < const NormalBasedSignatureEstimation < PointT, PointNT, PointFeature > > | ConstPtr |
typedef pcl::PointCloud < PointFeature > | FeatureCloud |
typedef boost::shared_ptr < NormalBasedSignatureEstimation < PointT, PointNT, PointFeature > > | Ptr |
Public Member Functions | |
size_t | getM () |
Returns the M parameter - the length of the rows used for the Discrete Cosine Transform. | |
size_t | getMPrime () |
Returns the M' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector. | |
size_t | getN () |
Returns the N parameter - the length of the columns used for the Discrete Fourier Transform. | |
size_t | getNPrime () |
Returns the N' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector. | |
float | getScale () |
Returns the scale parameter - used to determine the radius of the sampling disc around the point of interest - linked to the smoothing scale of the input cloud. | |
NormalBasedSignatureEstimation () | |
Empty constructor, initializes the internal parameters to the default values. | |
void | setM (size_t m) |
Setter method for the M parameter - the length of the rows used for the Discrete Cosine Transform. | |
void | setMPrime (size_t m_prime) |
Setter method for the M' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector. | |
void | setN (size_t n) |
Setter method for the N parameter - the length of the columns used for the Discrete Fourier Transform. | |
void | setNPrime (size_t n_prime) |
Setter method for the N' parameter - the number of columns to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector. | |
void | setScale (float scale) |
Setter method for the scale parameter - used to determine the radius of the sampling disc around the point of interest - linked to the smoothing scale of the input cloud. | |
Protected Member Functions | |
void | computeFeature (FeatureCloud &output) |
Abstract feature estimation method. | |
Private Member Functions | |
void | computeFeatureEigen (pcl::PointCloud< Eigen::MatrixXf > &) |
Make the computeFeature (&Eigen::MatrixXf); inaccessible from outside the class. | |
Private Attributes | |
size_t | M_ |
size_t | M_prime_ |
size_t | N_ |
size_t | N_prime_ |
float | scale_h_ |
Normal-based feature signature estimation class. Obtains the feature vector by applying Discrete Cosine and Fourier Transforms on an NxM array of real numbers representing the projection distances of the points in the input cloud to a disc around the point of interest. Please consult the following publication for more details: Xinju Li and Igor Guskov Multi-scale features for approximate alignment of point-based surfaces Proceedings of the third Eurographics symposium on Geometry processing July 2005, Vienna, Austria.
Definition at line 59 of file normal_based_signature.h.
typedef boost::shared_ptr<const NormalBasedSignatureEstimation<PointT, PointNT, PointFeature> > pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::ConstPtr |
Reimplemented from pcl::FeatureFromNormals< PointT, PointNT, PointFeature >.
Definition at line 70 of file normal_based_signature.h.
typedef pcl::PointCloud<PointFeature> pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::FeatureCloud |
Definition at line 68 of file normal_based_signature.h.
typedef boost::shared_ptr<NormalBasedSignatureEstimation<PointT, PointNT, PointFeature> > pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::Ptr |
Reimplemented from pcl::FeatureFromNormals< PointT, PointNT, PointFeature >.
Definition at line 69 of file normal_based_signature.h.
pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::NormalBasedSignatureEstimation | ( | ) | [inline] |
Empty constructor, initializes the internal parameters to the default values.
Definition at line 76 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::computeFeature | ( | FeatureCloud & | output | ) | [protected, virtual] |
Abstract feature estimation method.
[out] | output | the resultant features |
Implements pcl::Feature< PointT, PointFeature >.
Definition at line 44 of file normal_based_signature.hpp.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::computeFeatureEigen | ( | pcl::PointCloud< Eigen::MatrixXf > & | ) | [inline, private, virtual] |
Make the computeFeature (&Eigen::MatrixXf); inaccessible from outside the class.
[out] | output | the output point cloud |
Implements pcl::Feature< PointT, PointFeature >.
Definition at line 161 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::getM | ( | ) | [inline] |
Returns the M parameter - the length of the rows used for the Discrete Cosine Transform.
Definition at line 104 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::getMPrime | ( | ) | [inline] |
Returns the M' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector.
Definition at line 134 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::getN | ( | ) | [inline] |
Returns the N parameter - the length of the columns used for the Discrete Fourier Transform.
Definition at line 94 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::getNPrime | ( | ) | [inline] |
Returns the N' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector.
Definition at line 119 of file normal_based_signature.h.
float pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::getScale | ( | ) | [inline] |
Returns the scale parameter - used to determine the radius of the sampling disc around the point of interest - linked to the smoothing scale of the input cloud.
Definition at line 146 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::setM | ( | size_t | m | ) | [inline] |
Setter method for the M parameter - the length of the rows used for the Discrete Cosine Transform.
[in] | m | the length of the rows used for the Discrete Cosine Transform. |
Definition at line 100 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::setMPrime | ( | size_t | m_prime | ) | [inline] |
Setter method for the M' parameter - the number of rows to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector.
[in] | m_prime | the number of rows from the matrix of DFT and DCT that will be contained in the output |
Definition at line 127 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::setN | ( | size_t | n | ) | [inline] |
Setter method for the N parameter - the length of the columns used for the Discrete Fourier Transform.
[in] | n | the length of the columns used for the Discrete Fourier Transform. |
Definition at line 90 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::setNPrime | ( | size_t | n_prime | ) | [inline] |
Setter method for the N' parameter - the number of columns to be taken from the matrix of DFT and DCT values that will be contained in the output feature vector.
[in] | n_prime | the number of columns from the matrix of DFT and DCT that will be contained in the output |
Definition at line 112 of file normal_based_signature.h.
void pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::setScale | ( | float | scale | ) | [inline] |
Setter method for the scale parameter - used to determine the radius of the sampling disc around the point of interest - linked to the smoothing scale of the input cloud.
Definition at line 140 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::M_ [private] |
Definition at line 155 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::M_prime_ [private] |
Definition at line 155 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::N_ [private] |
Definition at line 155 of file normal_based_signature.h.
size_t pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::N_prime_ [private] |
Definition at line 155 of file normal_based_signature.h.
float pcl::NormalBasedSignatureEstimation< PointT, PointNT, PointFeature >::scale_h_ [private] |
Definition at line 154 of file normal_based_signature.h.