Point Cloud Library (PCL)  1.14.1-dev
sac_model_registration.h
1 /*
2  * Software License Agreement (BSD License)
3  *
4  * Point Cloud Library (PCL) - www.pointclouds.org
5  * Copyright (c) 2010-2011, 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/memory.h>
44 #include <pcl/pcl_macros.h>
45 #include <pcl/pcl_base.h>
46 #include <pcl/sample_consensus/sac_model.h>
47 #include <pcl/sample_consensus/model_types.h>
48 #include <pcl/common/eigen.h>
49 #include <pcl/common/centroid.h>
50 #include <map>
51 #include <numeric> // for std::iota
52 
53 namespace pcl
54 {
55  /** \brief SampleConsensusModelRegistration defines a model for Point-To-Point registration outlier rejection.
56  * \author Radu Bogdan Rusu
57  * \ingroup sample_consensus
58  */
59  template <typename PointT>
61  {
62  public:
68 
72 
73  using Ptr = shared_ptr<SampleConsensusModelRegistration<PointT> >;
74  using ConstPtr = shared_ptr<const SampleConsensusModelRegistration<PointT>>;
75 
76  /** \brief Constructor for base SampleConsensusModelRegistration.
77  * \param[in] cloud the input point cloud dataset
78  * \param[in] random if true set the random seed to the current time, else set to 12345 (default: false)
79  */
81  bool random = false)
82  : SampleConsensusModel<PointT> (cloud, random)
83  , target_ ()
85  {
86  // Call our own setInputCloud
87  setInputCloud (cloud);
88  model_name_ = "SampleConsensusModelRegistration";
89  sample_size_ = 3;
90  model_size_ = 16;
91  }
92 
93  /** \brief Constructor for base SampleConsensusModelRegistration.
94  * \param[in] cloud the input point cloud dataset
95  * \param[in] indices a vector of point indices to be used from \a cloud
96  * \param[in] random if true set the random seed to the current time, else set to 12345 (default: false)
97  */
99  const Indices &indices,
100  bool random = false)
101  : SampleConsensusModel<PointT> (cloud, indices, random)
102  , target_ ()
103  , sample_dist_thresh_ (0)
104  {
106  computeSampleDistanceThreshold (cloud, indices);
107  model_name_ = "SampleConsensusModelRegistration";
108  sample_size_ = 3;
109  model_size_ = 16;
110  }
111 
112  /** \brief Empty destructor */
113  ~SampleConsensusModelRegistration () override = default;
114 
115  /** \brief Provide a pointer to the input dataset
116  * \param[in] cloud the const boost shared pointer to a PointCloud message
117  */
118  inline void
119  setInputCloud (const PointCloudConstPtr &cloud) override
120  {
124  }
125 
126  /** \brief Set the input point cloud target.
127  * \param[in] target the input point cloud target
128  */
129  inline void
131  {
132  target_ = target;
133  // Cache the size and fill the target indices
134  const auto target_size = static_cast<index_t> (target->size ());
135  indices_tgt_.reset (new Indices (target_size));
136  std::iota (indices_tgt_->begin (), indices_tgt_->end (), 0);
138  }
139 
140  /** \brief Set the input point cloud target.
141  * \param[in] target the input point cloud target
142  * \param[in] indices_tgt a vector of point indices to be used from \a target
143  */
144  inline void
145  setInputTarget (const PointCloudConstPtr &target, const Indices &indices_tgt)
146  {
147  target_ = target;
148  indices_tgt_.reset (new Indices (indices_tgt));
150  }
151 
152  /** \brief Compute a 4x4 rigid transformation matrix from the samples given
153  * \param[in] samples the indices found as good candidates for creating a valid model
154  * \param[out] model_coefficients the resultant model coefficients
155  */
156  bool
157  computeModelCoefficients (const Indices &samples,
158  Eigen::VectorXf &model_coefficients) const override;
159 
160  /** \brief Compute all distances from the transformed points to their correspondences
161  * \param[in] model_coefficients the 4x4 transformation matrix
162  * \param[out] distances the resultant estimated distances
163  */
164  void
165  getDistancesToModel (const Eigen::VectorXf &model_coefficients,
166  std::vector<double> &distances) const override;
167 
168  /** \brief Select all the points which respect the given model coefficients as inliers.
169  * \param[in] model_coefficients the 4x4 transformation matrix
170  * \param[in] threshold a maximum admissible distance threshold for determining the inliers from the outliers
171  * \param[out] inliers the resultant model inliers
172  */
173  void
174  selectWithinDistance (const Eigen::VectorXf &model_coefficients,
175  const double threshold,
176  Indices &inliers) override;
177 
178  /** \brief Count all the points which respect the given model coefficients as inliers.
179  *
180  * \param[in] model_coefficients the coefficients of a model that we need to compute distances to
181  * \param[in] threshold maximum admissible distance threshold for determining the inliers from the outliers
182  * \return the resultant number of inliers
183  */
184  std::size_t
185  countWithinDistance (const Eigen::VectorXf &model_coefficients,
186  const double threshold) const override;
187 
188  /** \brief Recompute the 4x4 transformation using the given inlier set
189  * \param[in] inliers the data inliers found as supporting the model
190  * \param[in] model_coefficients the initial guess for the optimization
191  * \param[out] optimized_coefficients the resultant recomputed transformation
192  */
193  void
194  optimizeModelCoefficients (const Indices &inliers,
195  const Eigen::VectorXf &model_coefficients,
196  Eigen::VectorXf &optimized_coefficients) const override;
197 
198  void
200  const Eigen::VectorXf &,
201  PointCloud &, bool = true) const override
202  {
203  };
204 
205  bool
206  doSamplesVerifyModel (const std::set<index_t> &,
207  const Eigen::VectorXf &,
208  const double) const override
209  {
210  return (false);
211  }
212 
213  /** \brief Return a unique id for this model (SACMODEL_REGISTRATION). */
214  inline pcl::SacModel
215  getModelType () const override { return (SACMODEL_REGISTRATION); }
216 
217  protected:
220 
221  /** \brief Check if a sample of indices results in a good sample of points
222  * indices.
223  * \param[in] samples the resultant index samples
224  */
225  bool
226  isSampleGood (const Indices &samples) const override;
227 
228  /** \brief Computes an "optimal" sample distance threshold based on the
229  * principal directions of the input cloud.
230  * \param[in] cloud the const boost shared pointer to a PointCloud message
231  */
232  inline void
234  {
235  // Compute the principal directions via PCA
236  Eigen::Vector4f xyz_centroid;
237  Eigen::Matrix3f covariance_matrix;
238 
239  if (computeMeanAndCovarianceMatrix (*cloud, covariance_matrix, xyz_centroid) == 0) {
240  PCL_ERROR ("[pcl::SampleConsensusModelRegistration::computeSampleDistanceThreshold] No valid points in cloud!\n");
241  return;
242  }
243 
244  // Check if the covariance matrix is finite or not.
245  for (int i = 0; i < 3; ++i)
246  for (int j = 0; j < 3; ++j)
247  if (!std::isfinite (covariance_matrix.coeffRef (i, j)))
248  PCL_ERROR ("[pcl::SampleConsensusModelRegistration::computeSampleDistanceThreshold] Covariance matrix has NaN values! Is the input cloud finite?\n");
249 
250  Eigen::Vector3f eigen_values;
251  pcl::eigen33 (covariance_matrix, eigen_values);
252 
253  // Compute the distance threshold for sample selection
254  sample_dist_thresh_ = eigen_values.array ().sqrt ().sum () / 3.0;
256  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::setInputCloud] Estimated a sample selection distance threshold of: %f\n", sample_dist_thresh_);
257  }
258 
259  /** \brief Computes an "optimal" sample distance threshold based on the
260  * principal directions of the input cloud.
261  * \param[in] cloud the const boost shared pointer to a PointCloud message
262  * \param indices
263  */
264  inline void
266  const Indices &indices)
267  {
268  // Compute the principal directions via PCA
269  Eigen::Vector4f xyz_centroid;
270  Eigen::Matrix3f covariance_matrix;
271  if (computeMeanAndCovarianceMatrix (*cloud, indices, covariance_matrix, xyz_centroid) == 0) {
272  PCL_ERROR ("[pcl::SampleConsensusModelRegistration::computeSampleDistanceThreshold] No valid points given by cloud and indices!\n");
273  return;
274  }
275 
276  // Check if the covariance matrix is finite or not.
277  for (int i = 0; i < 3; ++i)
278  for (int j = 0; j < 3; ++j)
279  if (!std::isfinite (covariance_matrix.coeffRef (i, j)))
280  PCL_ERROR ("[pcl::SampleConsensusModelRegistration::computeSampleDistanceThreshold] Covariance matrix has NaN values! Is the input cloud finite?\n");
281 
282  Eigen::Vector3f eigen_values;
283  pcl::eigen33 (covariance_matrix, eigen_values);
284 
285  // Compute the distance threshold for sample selection
286  sample_dist_thresh_ = eigen_values.array ().sqrt ().sum () / 3.0;
288  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::setInputCloud] Estimated a sample selection distance threshold of: %f\n", sample_dist_thresh_);
289  }
290 
291  /** \brief Estimate a rigid transformation between a source and a target point cloud using an SVD closed-form
292  * solution of absolute orientation using unit quaternions
293  * \param[in] cloud_src the source point cloud dataset
294  * \param[in] indices_src the vector of indices describing the points of interest in cloud_src
295  * \param[in] cloud_tgt the target point cloud dataset
296  * \param[in] indices_tgt the vector of indices describing the correspondences of the interest points from
297  * indices_src
298  * \param[out] transform the resultant transformation matrix (as model coefficients)
299  *
300  * This method is an implementation of: Horn, B. “Closed-Form Solution of Absolute Orientation Using Unit Quaternions,” JOSA A, Vol. 4, No. 4, 1987
301  */
302  void
304  const Indices &indices_src,
305  const pcl::PointCloud<PointT> &cloud_tgt,
306  const Indices &indices_tgt,
307  Eigen::VectorXf &transform) const;
308 
309  /** \brief Compute mappings between original indices of the input_/target_ clouds. */
310  void
312  {
313  if (!indices_tgt_)
314  {
315  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::computeOriginalIndexMapping] Cannot compute mapping: indices_tgt_ is null.\n");
316  return;
317  }
318  if (!indices_)
319  {
320  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::computeOriginalIndexMapping] Cannot compute mapping: indices_ is null.\n");
321  return;
322  }
323  if (indices_->empty ())
324  {
325  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::computeOriginalIndexMapping] Cannot compute mapping: indices_ is empty.\n");
326  return;
327  }
328  if (indices_->size () != indices_tgt_->size ())
329  {
330  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::computeOriginalIndexMapping] Cannot compute mapping: indices_ and indices_tgt_ are not the same size (%zu vs %zu).\n",
331  indices_->size (), indices_tgt_->size ());
332  return;
333  }
334  for (std::size_t i = 0; i < indices_->size (); ++i)
335  correspondences_[(*indices_)[i]] = (*indices_tgt_)[i];
336  PCL_DEBUG ("[pcl::SampleConsensusModelRegistration::computeOriginalIndexMapping] Successfully computed mapping.\n");
337  }
338 
339  /** \brief A boost shared pointer to the target point cloud data array. */
341 
342  /** \brief A pointer to the vector of target point indices to use. */
344 
345  /** \brief Given the index in the original point cloud, give the matching original index in the target cloud */
346  std::map<index_t, index_t> correspondences_;
347 
348  /** \brief Internal distance threshold used for the sample selection step. */
350  public:
352  };
353 }
354 
355 #include <pcl/sample_consensus/impl/sac_model_registration.hpp>
Define methods for centroid estimation and covariance matrix calculus.
PointCloud represents the base class in PCL for storing collections of 3D points.
Definition: point_cloud.h:173
SampleConsensusModel represents the base model class.
Definition: sac_model.h:71
shared_ptr< SampleConsensusModel< PointT > > Ptr
Definition: sac_model.h:78
unsigned int sample_size_
The size of a sample from which the model is computed.
Definition: sac_model.h:589
typename PointCloud::ConstPtr PointCloudConstPtr
Definition: sac_model.h:74
IndicesPtr indices_
A pointer to the vector of point indices to use.
Definition: sac_model.h:557
virtual void setInputCloud(const PointCloudConstPtr &cloud)
Provide a pointer to the input dataset.
Definition: sac_model.h:301
std::string model_name_
The model name.
Definition: sac_model.h:551
unsigned int model_size_
The number of coefficients in the model.
Definition: sac_model.h:592
typename PointCloud::Ptr PointCloudPtr
Definition: sac_model.h:75
shared_ptr< const SampleConsensusModel< PointT > > ConstPtr
Definition: sac_model.h:79
SampleConsensusModelRegistration defines a model for Point-To-Point registration outlier rejection.
std::map< index_t, index_t > correspondences_
Given the index in the original point cloud, give the matching original index in the target cloud.
std::size_t countWithinDistance(const Eigen::VectorXf &model_coefficients, const double threshold) const override
Count all the points which respect the given model coefficients as inliers.
SampleConsensusModelRegistration(const PointCloudConstPtr &cloud, const Indices &indices, bool random=false)
Constructor for base SampleConsensusModelRegistration.
IndicesPtr indices_tgt_
A pointer to the vector of target point indices to use.
bool computeModelCoefficients(const Indices &samples, Eigen::VectorXf &model_coefficients) const override
Compute a 4x4 rigid transformation matrix from the samples given.
typename SampleConsensusModel< PointT >::PointCloudConstPtr PointCloudConstPtr
PointCloudConstPtr target_
A boost shared pointer to the target point cloud data array.
pcl::SacModel getModelType() const override
Return a unique id for this model (SACMODEL_REGISTRATION).
void getDistancesToModel(const Eigen::VectorXf &model_coefficients, std::vector< double > &distances) const override
Compute all distances from the transformed points to their correspondences.
void projectPoints(const Indices &, const Eigen::VectorXf &, PointCloud &, bool=true) const override
Create a new point cloud with inliers projected onto the model.
~SampleConsensusModelRegistration() override=default
Empty destructor.
void setInputTarget(const PointCloudConstPtr &target, const Indices &indices_tgt)
Set the input point cloud target.
void setInputTarget(const PointCloudConstPtr &target)
Set the input point cloud target.
SampleConsensusModelRegistration(const PointCloudConstPtr &cloud, bool random=false)
Constructor for base SampleConsensusModelRegistration.
bool doSamplesVerifyModel(const std::set< index_t > &, const Eigen::VectorXf &, const double) const override
Verify whether a subset of indices verifies a given set of model coefficients.
void estimateRigidTransformationSVD(const pcl::PointCloud< PointT > &cloud_src, const Indices &indices_src, const pcl::PointCloud< PointT > &cloud_tgt, const Indices &indices_tgt, Eigen::VectorXf &transform) const
Estimate a rigid transformation between a source and a target point cloud using an SVD closed-form so...
void computeSampleDistanceThreshold(const PointCloudConstPtr &cloud, const Indices &indices)
Computes an "optimal" sample distance threshold based on the principal directions of the input cloud.
bool isSampleGood(const Indices &samples) const override
Check if a sample of indices results in a good sample of points indices.
void optimizeModelCoefficients(const Indices &inliers, const Eigen::VectorXf &model_coefficients, Eigen::VectorXf &optimized_coefficients) const override
Recompute the 4x4 transformation using the given inlier set.
double sample_dist_thresh_
Internal distance threshold used for the sample selection step.
void computeSampleDistanceThreshold(const PointCloudConstPtr &cloud)
Computes an "optimal" sample distance threshold based on the principal directions of the input cloud.
void setInputCloud(const PointCloudConstPtr &cloud) override
Provide a pointer to the input dataset.
void computeOriginalIndexMapping()
Compute mappings between original indices of the input_/target_ clouds.
void selectWithinDistance(const Eigen::VectorXf &model_coefficients, const double threshold, Indices &inliers) override
Select all the points which respect the given model coefficients as inliers.
#define PCL_MAKE_ALIGNED_OPERATOR_NEW
Macro to signal a class requires a custom allocator.
Definition: memory.h:63
unsigned int computeMeanAndCovarianceMatrix(const pcl::PointCloud< PointT > &cloud, Eigen::Matrix< Scalar, 3, 3 > &covariance_matrix, Eigen::Matrix< Scalar, 4, 1 > &centroid)
Compute the normalized 3x3 covariance matrix and the centroid of a given set of points in a single lo...
Definition: centroid.hpp:509
void eigen33(const Matrix &mat, typename Matrix::Scalar &eigenvalue, Vector &eigenvector)
determines the eigenvector and eigenvalue of the smallest eigenvalue of the symmetric positive semi d...
Definition: eigen.hpp:295
Defines functions, macros and traits for allocating and using memory.
detail::int_type_t< detail::index_type_size, detail::index_type_signed > index_t
Type used for an index in PCL.
Definition: types.h:112
SacModel
Definition: model_types.h:46
@ SACMODEL_REGISTRATION
Definition: model_types.h:60
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition: types.h:133
shared_ptr< Indices > IndicesPtr
Definition: pcl_base.h:58
Defines all the PCL and non-PCL macros used.
A point structure representing Euclidean xyz coordinates, and the RGB color.