Main MRPT website > C++ reference for MRPT 1.4.0
model_search.h
Go to the documentation of this file.
1/* +---------------------------------------------------------------------------+
2 | Mobile Robot Programming Toolkit (MRPT) |
3 | http://www.mrpt.org/ |
4 | |
5 | Copyright (c) 2005-2016, Individual contributors, see AUTHORS file |
6 | See: http://www.mrpt.org/Authors - All rights reserved. |
7 | Released under BSD License. See details in http://www.mrpt.org/License |
8 +---------------------------------------------------------------------------+ */
9
10#ifndef math_modelsearch_h
11#define math_modelsearch_h
12
14#include <set>
15
16namespace mrpt {
17 namespace math {
18
19
20 /** Model search implementations: RANSAC and genetic algorithm
21 *
22 * The type \a TModelFit is a user-supplied struct/class that implements this interface:
23 * - Types:
24 * - \a Real : The numeric type to use (typ: double, float)
25 * - \a Model : The type of the model to be fitted (for example: A matrix, a TLine2D, a TPlane3D, ...)
26 * - Methods:
27 * - size_t getSampleCount() const : return the number of samples. This should not change during a model search.
28 * - bool fitModel( const vector_size_t& useIndices, Model& model ) const : This function fits a model to the data selected by the indices. The return value indicates the success, hence false means a degenerate case, where no model was found.
29 * - Real testSample( size_t index, const Model& model ) const : return some value that indicates how well a sample fits to the model. This way the thresholding is moved to the searching procedure and the model just tells how good a sample is.
30 *
31 * There are two methods provided in this class to fit a model:
32 * - \a ransacSingleModel (RANSAC): Just like mrpt::math::RANSAC_Template
33 *
34 * - \a geneticSingleModel (Genetic): Provides a mixture of a genetic and the ransac algorithm.
35 * Instead of selecting a set of data in each iteration, it takes more (ex. 10) and order these model
36 * using some fitness function: the average error of the inliers scaled by the number of outliers (This
37 * fitness might require some fine tuning). Than the (ex 10) new kernel for the next iteration is created as follows:
38 * - Take the best kernels (as for the original ransac)
39 * - Select two kernels ( with a higher probability for the better models) and let the new kernel be a subset of the two original kernels ( additionally to leave the local minimums an additional random seed might appear - mutation)
40 * - Generate some new random samples.
41 *
42 * For an example of usage, see "samples/model_search_test/"
43 * \sa mrpt::math::RANSAC_Template, another RANSAC implementation where models can be matrices only.
44 *
45 * \author Zoltar Gaal
46 * \ingroup ransac_grp
47 */
49 private:
50 //! Select random (unique) indices from the 0..p_size sequence
51 void pickRandomIndex( size_t p_size, size_t p_pick, vector_size_t& p_ind );
52
53 /** Select random (unique) indices from the set.
54 * The set is destroyed during pick */
55 void pickRandomIndex( std::set<size_t> p_set, size_t p_pick, vector_size_t& p_ind );
56
57 public:
58 template<typename TModelFit>
59 bool ransacSingleModel( const TModelFit& p_state,
60 size_t p_kernelSize,
61 const typename TModelFit::Real& p_fitnessThreshold,
62 typename TModelFit::Model& p_bestModel,
63 vector_size_t& p_inliers );
64
65 private:
66 template<typename TModelFit>
67 struct TSpecies {
68 typename TModelFit::Model model;
71 typename TModelFit::Real fitness;
72
73 static bool compare( const TSpecies* p_a, const TSpecies* p_b )
74 {
75 return p_a->fitness < p_b->fitness;
76 }
77 };
78
79 public:
80 template<typename TModelFit>
81 bool geneticSingleModel( const TModelFit& p_state,
82 size_t p_kernelSize,
83 const typename TModelFit::Real& p_fitnessThreshold,
84 size_t p_populationSize,
85 size_t p_maxIteration,
86 typename TModelFit::Model& p_bestModel,
87 vector_size_t& p_inliers );
88 }; // end of class
89
90 } // namespace math
91} // namespace mrpt
92
93// Template implementations:
94#include "model_search_impl.h"
95
96#endif // math_modelsearch_h
Model search implementations: RANSAC and genetic algorithm.
void pickRandomIndex(size_t p_size, size_t p_pick, vector_size_t &p_ind)
Select random (unique) indices from the 0..p_size sequence.
void pickRandomIndex(std::set< size_t > p_set, size_t p_pick, vector_size_t &p_ind)
Select random (unique) indices from the set.
std::vector< size_t > vector_size_t
This is the global namespace for all Mobile Robot Programming Toolkit (MRPT) libraries.
static bool compare(const TSpecies *p_a, const TSpecies *p_b)



Page generated by Doxygen 1.9.7 for MRPT 1.4.0 SVN: at Tue Jun 27 15:23:24 UTC 2023