Point Cloud Library (PCL)  1.14.1-dev
sample_consensus_prerejective.h
1 /*
2  * Software License Agreement (BSD License)
3  *
4  * Point Cloud Library (PCL) - www.pointclouds.org
5  * Copyright (c) 2010-2012, Willow Garage, Inc.
6  * Copyright (c) 2012-, Open Perception, Inc.
7  *
8  * All rights reserved.
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  *
14  * * Redistributions of source code must retain the above copyright
15  * notice, this list of conditions and the following disclaimer.
16  * * Redistributions in binary form must reproduce the above
17  * copyright notice, this list of conditions and the following
18  * disclaimer in the documentation and/or other materials provided
19  * with the distribution.
20  * * Neither the name of the copyright holder(s) nor the names of its
21  * contributors may be used to endorse or promote products derived
22  * from this software without specific prior written permission.
23  *
24  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
25  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
26  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
27  * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
28  * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
29  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
30  * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
31  * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
32  * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
34  * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
35  * POSSIBILITY OF SUCH DAMAGE.
36  *
37  * $Id$
38  *
39  */
40 
41 #pragma once
42 
43 #include <pcl/registration/correspondence_rejection_poly.h>
44 #include <pcl/registration/registration.h>
45 #include <pcl/registration/transformation_estimation_svd.h>
46 #include <pcl/registration/transformation_validation.h>
47 
48 namespace pcl {
49 /** \brief Pose estimation and alignment class using a prerejective RANSAC routine.
50  *
51  * This class inserts a simple, yet effective "prerejection" step into the standard
52  * RANSAC pose estimation loop in order to avoid verification of pose hypotheses
53  * that are likely to be wrong. This is achieved by local pose-invariant geometric
54  * constraints, as also implemented in the class
55  * \ref registration::CorrespondenceRejectorPoly "CorrespondenceRejectorPoly".
56  *
57  * In order to robustly align partial/occluded models, this routine performs
58  * fit error evaluation using only inliers, i.e. points closer than a
59  * Euclidean threshold, which is specifiable using \ref setInlierFraction().
60  *
61  * The amount of prerejection or "greedyness" of the algorithm can be specified
62  * using \ref setSimilarityThreshold() in [0,1[, where a value of 0 means disabled,
63  * and 1 is maximally rejective.
64  *
65  * If you use this in academic work, please cite:
66  *
67  * A. G. Buch, D. Kraft, J.-K. Kämäräinen, H. G. Petersen and N. Krüger.
68  * Pose Estimation using Local Structure-Specific Shape and Appearance Context.
69  * International Conference on Robotics and Automation (ICRA), 2013.
70  *
71  * \author Anders Glent Buch (andersgb1@gmail.com)
72  * \ingroup registration
73  */
74 template <typename PointSource, typename PointTarget, typename FeatureT>
75 class SampleConsensusPrerejective : public Registration<PointSource, PointTarget> {
76 public:
78 
91 
96 
99 
102 
106 
107  using Ptr =
108  shared_ptr<SampleConsensusPrerejective<PointSource, PointTarget, FeatureT>>;
109  using ConstPtr =
110  shared_ptr<const SampleConsensusPrerejective<PointSource, PointTarget, FeatureT>>;
111 
113 
119 
120  /** \brief Constructor */
122  : input_features_()
123  , target_features_()
124  , feature_tree_(new pcl::KdTreeFLANN<FeatureT>)
126  {
127  reg_name_ = "SampleConsensusPrerejective";
128  correspondence_rejector_poly_->setSimilarityThreshold(0.6f);
129  max_iterations_ = 5000;
132  };
133 
134  /** \brief Destructor */
135  ~SampleConsensusPrerejective() override = default;
136 
137  /** \brief Provide a boost shared pointer to the source point cloud's feature
138  * descriptors \param features the source point cloud's features
139  */
140  void
141  setSourceFeatures(const FeatureCloudConstPtr& features);
142 
143  /** \brief Get a pointer to the source point cloud's features */
144  inline const FeatureCloudConstPtr
146  {
147  return (input_features_);
148  }
149 
150  /** \brief Provide a boost shared pointer to the target point cloud's feature
151  * descriptors \param features the target point cloud's features
152  */
153  void
154  setTargetFeatures(const FeatureCloudConstPtr& features);
155 
156  /** \brief Get a pointer to the target point cloud's features */
157  inline const FeatureCloudConstPtr
159  {
160  return (target_features_);
161  }
162 
163  /** \brief Set the number of samples to use during each iteration
164  * \param nr_samples the number of samples to use during each iteration
165  */
166  inline void
167  setNumberOfSamples(int nr_samples)
168  {
169  nr_samples_ = nr_samples;
170  }
171 
172  /** \brief Get the number of samples to use during each iteration, as set by the user
173  */
174  inline int
176  {
177  return (nr_samples_);
178  }
179 
180  /** \brief Set the number of neighbors to use when selecting a random feature
181  * correspondence. A higher value will add more randomness to the feature matching.
182  * \param k the number of neighbors to use when selecting a random feature
183  * correspondence.
184  */
185  inline void
187  {
188  k_correspondences_ = k;
189  }
190 
191  /** \brief Get the number of neighbors used when selecting a random feature
192  * correspondence, as set by the user */
193  inline int
195  {
196  return (k_correspondences_);
197  }
198 
199  /** \brief Set the similarity threshold in [0,1[ between edge lengths of the
200  * underlying polygonal correspondence rejector object, where 1 is a perfect match
201  * \param similarity_threshold edge length similarity threshold
202  */
203  inline void
204  setSimilarityThreshold(float similarity_threshold)
205  {
206  correspondence_rejector_poly_->setSimilarityThreshold(similarity_threshold);
207  }
208 
209  /** \brief Get the similarity threshold between edge lengths of the underlying
210  * polygonal correspondence rejector object, \return edge length similarity threshold
211  */
212  inline float
214  {
215  return correspondence_rejector_poly_->getSimilarityThreshold();
216  }
217 
218  /** \brief Set the required inlier fraction (of the input)
219  * \param inlier_fraction required inlier fraction, must be in [0,1]
220  */
221  inline void
222  setInlierFraction(float inlier_fraction)
223  {
224  inlier_fraction_ = inlier_fraction;
225  }
226 
227  /** \brief Get the required inlier fraction
228  * \return required inlier fraction in [0,1]
229  */
230  inline float
232  {
233  return inlier_fraction_;
234  }
235 
236  /** \brief Get the inlier indices of the source point cloud under the final
237  * transformation
238  * @return inlier indices
239  */
240  inline const pcl::Indices&
241  getInliers() const
242  {
243  return inliers_;
244  }
245 
246 protected:
247  /** \brief Choose a random index between 0 and n-1
248  * \param n the number of possible indices to choose from
249  */
250  inline int
251  getRandomIndex(int n) const
252  {
253  return (static_cast<int>(n * (rand() / (RAND_MAX + 1.0))));
254  };
255 
256  /** \brief Select \a nr_samples sample points from cloud while making sure that their
257  * pairwise distances are greater than a user-defined minimum distance, \a
258  * min_sample_distance. \param cloud the input point cloud \param nr_samples the
259  * number of samples to select \param sample_indices the resulting sample indices
260  */
261  void
262  selectSamples(const PointCloudSource& cloud,
263  int nr_samples,
264  pcl::Indices& sample_indices);
265 
266  /** \brief For each of the sample points, find a list of points in the target cloud
267  * whose features are similar to the sample points' features. From these, select one
268  * randomly which will be considered that sample point's correspondence. \param
269  * sample_indices the indices of each sample point \param similar_features
270  * correspondence cache, which is used to read/write already computed correspondences
271  * \param corresponding_indices the resulting indices of each sample's corresponding
272  * point in the target cloud
273  */
274  void
275  findSimilarFeatures(const pcl::Indices& sample_indices,
276  std::vector<pcl::Indices>& similar_features,
277  pcl::Indices& corresponding_indices);
278 
279  /** \brief Rigid transformation computation method.
280  * \param output the transformed input point cloud dataset using the rigid
281  * transformation found \param guess The computed transformation
282  */
283  void
285  const Eigen::Matrix4f& guess) override;
286 
287  /** \brief Obtain the fitness of a transformation
288  * The following metrics are calculated, based on
289  * \b final_transformation_ and \b corr_dist_threshold_:
290  * - Inliers: the number of transformed points which are closer than threshold to NN
291  * - Error score: the MSE of the inliers
292  * \param inliers indices of source point cloud inliers
293  * \param fitness_score output fitness score as RMSE
294  */
295  void
296  getFitness(pcl::Indices& inliers, float& fitness_score);
297 
298  /** \brief The source point cloud's feature descriptors. */
300 
301  /** \brief The target point cloud's feature descriptors. */
303 
304  /** \brief The number of samples to use during each iteration. */
305  int nr_samples_{3};
306 
307  /** \brief The number of neighbors to use when selecting a random feature
308  * correspondence. */
310 
311  /** \brief The KdTree used to compare feature descriptors. */
313 
314  /** \brief The polygonal correspondence rejector used for prerejection */
316 
317  /** \brief The fraction [0,1] of inlier points required for accepting a transformation
318  */
319  float inlier_fraction_{0.0f};
320 
321  /** \brief Inlier points of final transformation as indices into source */
323 };
324 } // namespace pcl
325 
326 #include <pcl/registration/impl/sample_consensus_prerejective.hpp>
KdTreeFLANN is a generic type of 3D spatial locator using kD-tree structures.
Definition: kdtree_flann.h:132
shared_ptr< PointCloud< PointSource > > Ptr
Definition: point_cloud.h:413
shared_ptr< const PointCloud< PointSource > > ConstPtr
Definition: point_cloud.h:414
Registration represents the base registration class for general purpose, ICP-like methods.
Definition: registration.h:57
std::string reg_name_
The registration method name.
Definition: registration.h:548
shared_ptr< const Registration< PointSource, PointTarget, Scalar > > ConstPtr
Definition: registration.h:67
typename PointCloudSource::ConstPtr PointCloudSourceConstPtr
Definition: registration.h:78
typename PointCloudSource::Ptr PointCloudSourcePtr
Definition: registration.h:77
shared_ptr< Registration< PointSource, PointTarget, Scalar > > Ptr
Definition: registration.h:66
Eigen::Matrix< Scalar, 4, 4 > Matrix4
Definition: registration.h:59
TransformationEstimationPtr transformation_estimation_
A TransformationEstimation object, used to calculate the 4x4 rigid transformation.
Definition: registration.h:625
int max_iterations_
The maximum number of iterations the internal optimization should run for.
Definition: registration.h:563
Pose estimation and alignment class using a prerejective RANSAC routine.
float getSimilarityThreshold() const
Get the similarity threshold between edge lengths of the underlying polygonal correspondence rejector...
float inlier_fraction_
The fraction [0,1] of inlier points required for accepting a transformation.
pcl::Indices inliers_
Inlier points of final transformation as indices into source.
float getInlierFraction() const
Get the required inlier fraction.
void setCorrespondenceRandomness(int k)
Set the number of neighbors to use when selecting a random feature correspondence.
const FeatureCloudConstPtr getSourceFeatures() const
Get a pointer to the source point cloud's features.
FeatureCloudConstPtr input_features_
The source point cloud's feature descriptors.
FeatureKdTreePtr feature_tree_
The KdTree used to compare feature descriptors.
void setTargetFeatures(const FeatureCloudConstPtr &features)
Provide a boost shared pointer to the target point cloud's feature descriptors.
void setInlierFraction(float inlier_fraction)
Set the required inlier fraction (of the input)
CorrespondenceRejectorPolyPtr correspondence_rejector_poly_
The polygonal correspondence rejector used for prerejection.
void computeTransformation(PointCloudSource &output, const Eigen::Matrix4f &guess) override
Rigid transformation computation method.
int getCorrespondenceRandomness() const
Get the number of neighbors used when selecting a random feature correspondence, as set by the user.
typename CorrespondenceRejectorPoly::ConstPtr CorrespondenceRejectorPolyConstPtr
const FeatureCloudConstPtr getTargetFeatures() const
Get a pointer to the target point cloud's features.
typename Registration< PointSource, PointTarget >::PointCloudSource PointCloudSource
typename FeatureCloud::ConstPtr FeatureCloudConstPtr
const pcl::Indices & getInliers() const
Get the inlier indices of the source point cloud under the final transformation.
void setSimilarityThreshold(float similarity_threshold)
Set the similarity threshold in [0,1[ between edge lengths of the underlying polygonal correspondence...
void findSimilarFeatures(const pcl::Indices &sample_indices, std::vector< pcl::Indices > &similar_features, pcl::Indices &corresponding_indices)
For each of the sample points, find a list of points in the target cloud whose features are similar t...
void setSourceFeatures(const FeatureCloudConstPtr &features)
Provide a boost shared pointer to the source point cloud's feature descriptors.
FeatureCloudConstPtr target_features_
The target point cloud's feature descriptors.
void setNumberOfSamples(int nr_samples)
Set the number of samples to use during each iteration.
typename KdTreeFLANN< FeatureT >::Ptr FeatureKdTreePtr
int getNumberOfSamples() const
Get the number of samples to use during each iteration, as set by the user.
~SampleConsensusPrerejective() override=default
Destructor.
void selectSamples(const PointCloudSource &cloud, int nr_samples, pcl::Indices &sample_indices)
Select nr_samples sample points from cloud while making sure that their pairwise distances are greate...
int getRandomIndex(int n) const
Choose a random index between 0 and n-1.
int nr_samples_
The number of samples to use during each iteration.
typename CorrespondenceRejectorPoly::Ptr CorrespondenceRejectorPolyPtr
int k_correspondences_
The number of neighbors to use when selecting a random feature correspondence.
void getFitness(pcl::Indices &inliers, float &fitness_score)
Obtain the fitness of a transformation The following metrics are calculated, based on final_transform...
CorrespondenceRejectorPoly implements a correspondence rejection method that exploits low-level and p...
shared_ptr< const CorrespondenceRejectorPoly< SourceT, TargetT > > ConstPtr
shared_ptr< CorrespondenceRejectorPoly< SourceT, TargetT > > Ptr
TransformationEstimationSVD implements SVD-based estimation of the transformation aligning the given ...
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition: types.h:133
shared_ptr< ::pcl::PointIndices > Ptr
Definition: PointIndices.h:13
shared_ptr< const ::pcl::PointIndices > ConstPtr
Definition: PointIndices.h:14