MRPT  1.9.9
mrpt::tfest::TMatchingPairList Class Reference

Detailed Description

A list of TMatchingPair.

Definition at line 81 of file TMatchingPair.h.

#include <mrpt/tfest/TMatchingPair.h>

Inheritance diagram for mrpt::tfest::TMatchingPairList:
Inheritance graph

Public Member Functions

bool indexOtherMapHasCorrespondence (size_t idx) const
 Checks if the given index from the "other" map appears in the list. More...
 
void dumpToFile (const std::string &fileName) const
 Saves the correspondences to a text file. More...
 
void saveAsMATLABScript (const std::string &filName) const
 Saves the correspondences as a MATLAB script which draws them. More...
 
float overallSquareError (const mrpt::poses::CPose2D &q) const
 Computes the overall square error between the 2D points in the list of correspondences, given the 2D transformation "q"

\[ \sum\limits_i e_i \]

Where $ e_i $ are the elements of the square error vector as computed by computeSquareErrorVector. More...

 
float overallSquareErrorAndPoints (const mrpt::poses::CPose2D &q, std::vector< float > &xs, std::vector< float > &ys) const
 Computes the overall square error between the 2D points in the list of correspondences, given the 2D transformation "q", and return the transformed points as well. More...
 
void squareErrorVector (const mrpt::poses::CPose2D &q, std::vector< float > &out_sqErrs) const
 Returns a vector with the square error between each pair of correspondences in the list, given the 2D transformation "q" Each element $ e_i $ is the square distance between the "this" (global) point and the "other" (local) point transformed through "q":

\[ e_i = | x_{this} - q \oplus x_{other} |^2 \]

. More...

 
void squareErrorVector (const mrpt::poses::CPose2D &q, std::vector< float > &out_sqErrs, std::vector< float > &xs, std::vector< float > &ys) const
 Returns a vector with the square error between each pair of correspondences in the list and the transformed "other" (local) points, given the 2D transformation "q" Each element $ e_i $ is the square distance between the "this" (global) point and the "other" (local) point transformed through "q":

\[ e_i = | x_{this} - q \oplus x_{other} |^2 \]

. More...

 
bool contains (const TMatchingPair &p) const
 Test whether the given pair "p" is within the pairings. More...
 
void filterUniqueRobustPairs (const size_t num_elements_this_map, TMatchingPairList &out_filtered_list) const
 Creates a filtered list of pairings with those ones which have a single correspondence which coincides in both directions, i.e. More...
 

Public Attributes

elements
 STL member. More...
 

Member Function Documentation

◆ contains()

bool TMatchingPairList::contains ( const TMatchingPair p) const

Test whether the given pair "p" is within the pairings.

Definition at line 122 of file TMatchingPair.cpp.

References begin().

◆ dumpToFile()

void TMatchingPairList::dumpToFile ( const std::string fileName) const

Saves the correspondences to a text file.

Definition at line 28 of file TMatchingPair.cpp.

References ASSERT_, begin(), and mrpt::format().

Referenced by mrpt::tfest::se2_l2_robust().

◆ filterUniqueRobustPairs()

void TMatchingPairList::filterUniqueRobustPairs ( const size_t  num_elements_this_map,
TMatchingPairList out_filtered_list 
) const

Creates a filtered list of pairings with those ones which have a single correspondence which coincides in both directions, i.e.

the best pairing of element i in map this is the best match for element j in map other, and viceversa

Definition at line 184 of file TMatchingPair.cpp.

Referenced by mrpt::maps::CPointsMap::determineMatching2D(), and mrpt::maps::CPointsMap::determineMatching3D().

◆ indexOtherMapHasCorrespondence()

bool TMatchingPairList::indexOtherMapHasCorrespondence ( size_t  idx) const

Checks if the given index from the "other" map appears in the list.

Definition at line 72 of file TMatchingPair.cpp.

References begin().

◆ overallSquareError()

float TMatchingPairList::overallSquareError ( const mrpt::poses::CPose2D q) const

Computes the overall square error between the 2D points in the list of correspondences, given the 2D transformation "q"

\[ \sum\limits_i e_i \]

Where $ e_i $ are the elements of the square error vector as computed by computeSquareErrorVector.

See also
squareErrorVector, overallSquareErrorAndPoints

Definition at line 104 of file TMatchingPair.cpp.

References mrpt::math::size().

◆ overallSquareErrorAndPoints()

float TMatchingPairList::overallSquareErrorAndPoints ( const mrpt::poses::CPose2D q,
std::vector< float > &  xs,
std::vector< float > &  ys 
) const

Computes the overall square error between the 2D points in the list of correspondences, given the 2D transformation "q", and return the transformed points as well.

\[ \sum\limits_i e_i \]

Where $ e_i $ are the elements of the square error vector as computed by computeSquareErrorVector

See also
squareErrorVector

Definition at line 111 of file TMatchingPair.cpp.

References mrpt::math::size().

◆ saveAsMATLABScript()

void TMatchingPairList::saveAsMATLABScript ( const std::string filName) const

Saves the correspondences as a MATLAB script which draws them.

Definition at line 41 of file TMatchingPair.cpp.

References begin(), mrpt::system::os::fclose(), mrpt::system::os::fopen(), and mrpt::system::os::fprintf().

◆ squareErrorVector() [1/2]

void TMatchingPairList::squareErrorVector ( const mrpt::poses::CPose2D q,
std::vector< float > &  out_sqErrs 
) const

Returns a vector with the square error between each pair of correspondences in the list, given the 2D transformation "q" Each element $ e_i $ is the square distance between the "this" (global) point and the "other" (local) point transformed through "q":

\[ e_i = | x_{this} - q \oplus x_{other} |^2 \]

.

See also
overallSquareError

Definition at line 132 of file TMatchingPair.cpp.

References begin(), mrpt::math::size(), and mrpt::square().

Referenced by mrpt::slam::CICP::ICP_Method_LM().

◆ squareErrorVector() [2/2]

void TMatchingPairList::squareErrorVector ( const mrpt::poses::CPose2D q,
std::vector< float > &  out_sqErrs,
std::vector< float > &  xs,
std::vector< float > &  ys 
) const

Returns a vector with the square error between each pair of correspondences in the list and the transformed "other" (local) points, given the 2D transformation "q" Each element $ e_i $ is the square distance between the "this" (global) point and the "other" (local) point transformed through "q":

\[ e_i = | x_{this} - q \oplus x_{other} |^2 \]

.

See also
overallSquareError

Definition at line 157 of file TMatchingPair.cpp.

References begin(), mrpt::math::size(), and mrpt::square().

Member Data Documentation

◆ elements

T std::vector< T >::elements
inherited

STL member.




Page generated by Doxygen 1.8.14 for MRPT 1.9.9 Git: 7d5e6d718 Fri Aug 24 01:51:28 2018 +0200 at lun nov 2 08:35:50 CET 2020