Point Cloud Library (PCL)  1.12.0
mls.h
1 /*
2  * Software License Agreement (BSD License)
3  *
4  * Point Cloud Library (PCL) - www.pointclouds.org
5  * Copyright (c) 2009-2011, Willow Garage, Inc.
6  *
7  * All rights reserved.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  *
13  * * Redistributions of source code must retain the above copyright
14  * notice, this list of conditions and the following disclaimer.
15  * * Redistributions in binary form must reproduce the above
16  * copyright notice, this list of conditions and the following
17  * disclaimer in the documentation and/or other materials provided
18  * with the distribution.
19  * * Neither the name of Willow Garage, Inc. nor the names of its
20  * contributors may be used to endorse or promote products derived
21  * from this software without specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
24  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
25  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
26  * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
27  * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
28  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
29  * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
30  * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
31  * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
33  * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34  * POSSIBILITY OF SUCH DAMAGE.
35  *
36  * $Id$
37  *
38  */
39 
40 #pragma once
41 
42 #include <functional>
43 #include <map>
44 #include <random>
45 #include <Eigen/Core> // for Vector3i, Vector3d, ...
46 
47 // PCL includes
48 #include <pcl/memory.h>
49 #include <pcl/pcl_base.h>
50 #include <pcl/pcl_macros.h>
51 #include <pcl/search/search.h> // for Search
52 
53 #include <pcl/surface/processing.h>
54 
55 namespace pcl
56 {
57 
58  /** \brief Data structure used to store the results of the MLS fitting */
59  struct MLSResult
60  {
62  {
63  NONE, /**< \brief Project to the mls plane. */
64  SIMPLE, /**< \brief Project along the mls plane normal to the polynomial surface. */
65  ORTHOGONAL /**< \brief Project to the closest point on the polynonomial surface. */
66  };
67 
68  /** \brief Data structure used to store the MLS polynomial partial derivatives */
70  {
71  double z; /**< \brief The z component of the polynomial evaluated at z(u, v). */
72  double z_u; /**< \brief The partial derivative dz/du. */
73  double z_v; /**< \brief The partial derivative dz/dv. */
74  double z_uu; /**< \brief The partial derivative d^2z/du^2. */
75  double z_vv; /**< \brief The partial derivative d^2z/dv^2. */
76  double z_uv; /**< \brief The partial derivative d^2z/dudv. */
77  };
78 
79  /** \brief Data structure used to store the MLS projection results */
81  {
82  MLSProjectionResults () : u (0), v (0) {}
83 
84  double u; /**< \brief The u-coordinate of the projected point in local MLS frame. */
85  double v; /**< \brief The v-coordinate of the projected point in local MLS frame. */
86  Eigen::Vector3d point; /**< \brief The projected point. */
87  Eigen::Vector3d normal; /**< \brief The projected point's normal. */
89  };
90 
91  inline
92  MLSResult () : num_neighbors (0), curvature (0.0f), order (0), valid (false) {}
93 
94  inline
95  MLSResult (const Eigen::Vector3d &a_query_point,
96  const Eigen::Vector3d &a_mean,
97  const Eigen::Vector3d &a_plane_normal,
98  const Eigen::Vector3d &a_u,
99  const Eigen::Vector3d &a_v,
100  const Eigen::VectorXd &a_c_vec,
101  const int a_num_neighbors,
102  const float a_curvature,
103  const int a_order);
104 
105  /** \brief Given a point calculate its 3D location in the MLS frame.
106  * \param[in] pt The point
107  * \param[out] u The u-coordinate of the point in local MLS frame.
108  * \param[out] v The v-coordinate of the point in local MLS frame.
109  * \param[out] w The w-coordinate of the point in local MLS frame.
110  */
111  inline void
112  getMLSCoordinates (const Eigen::Vector3d &pt, double &u, double &v, double &w) const;
113 
114  /** \brief Given a point calculate its 2D location in the MLS frame.
115  * \param[in] pt The point
116  * \param[out] u The u-coordinate of the point in local MLS frame.
117  * \param[out] v The v-coordinate of the point in local MLS frame.
118  */
119  inline void
120  getMLSCoordinates (const Eigen::Vector3d &pt, double &u, double &v) const;
121 
122  /** \brief Calculate the polynomial
123  * \param[in] u The u-coordinate of the point in local MLS frame.
124  * \param[in] v The v-coordinate of the point in local MLS frame.
125  * \return The polynomial value at the provided uv coordinates.
126  */
127  inline double
128  getPolynomialValue (const double u, const double v) const;
129 
130  /** \brief Calculate the polynomial's first and second partial derivatives.
131  * \param[in] u The u-coordinate of the point in local MLS frame.
132  * \param[in] v The v-coordinate of the point in local MLS frame.
133  * \return The polynomial partial derivatives at the provided uv coordinates.
134  */
135  inline PolynomialPartialDerivative
136  getPolynomialPartialDerivative (const double u, const double v) const;
137 
138  /** \brief Calculate the principal curvatures using the polynomial surface.
139  * \param[in] u The u-coordinate of the point in local MLS frame.
140  * \param[in] v The v-coordinate of the point in local MLS frame.
141  * \return The principal curvature [k1, k2] at the provided uv coordinates.
142  * \note If an error occurs then 1e-5 is returned.
143  */
144  Eigen::Vector2f
145  calculatePrincipalCurvatures (const double u, const double v) const;
146 
147  /** \brief Calculate the principal curvatures using the polynomial surface.
148  * \param[in] u The u-coordinate of the point in local MLS frame.
149  * \param[in] v The v-coordinate of the point in local MLS frame.
150  * \return The principal curvature [k1, k2] at the provided ub coordinates.
151  * \note If an error occurs then 1e-5 is returned.
152  */
153  PCL_DEPRECATED(1, 15, "use calculatePrincipalCurvatures() instead")
154  inline Eigen::Vector2f
155  calculatePrincipleCurvatures (const double u, const double v) const { return calculatePrincipalCurvatures(u, v); };
156 
157  /** \brief Project a point orthogonal to the polynomial surface.
158  * \param[in] u The u-coordinate of the point in local MLS frame.
159  * \param[in] v The v-coordinate of the point in local MLS frame.
160  * \param[in] w The w-coordinate of the point in local MLS frame.
161  * \return The MLSProjectionResults for the input data.
162  * \note If the MLSResults does not contain polynomial data it projects the point onto the mls plane.
163  * \note If the optimization diverges it performs a simple projection on to the polynomial surface.
164  * \note This was implemented based on this https://math.stackexchange.com/questions/1497093/shortest-distance-between-point-and-surface
165  */
166  inline MLSProjectionResults
167  projectPointOrthogonalToPolynomialSurface (const double u, const double v, const double w) const;
168 
169  /** \brief Project a point onto the MLS plane.
170  * \param[in] u The u-coordinate of the point in local MLS frame.
171  * \param[in] v The v-coordinate of the point in local MLS frame.
172  * \return The MLSProjectionResults for the input data.
173  */
174  inline MLSProjectionResults
175  projectPointToMLSPlane (const double u, const double v) const;
176 
177  /** \brief Project a point along the MLS plane normal to the polynomial surface.
178  * \param[in] u The u-coordinate of the point in local MLS frame.
179  * \param[in] v The v-coordinate of the point in local MLS frame.
180  * \return The MLSProjectionResults for the input data.
181  * \note If the MLSResults does not contain polynomial data it projects the point onto the mls plane.
182  */
183  inline MLSProjectionResults
184  projectPointSimpleToPolynomialSurface (const double u, const double v) const;
185 
186  /**
187  * \brief Project a point using the specified method.
188  * \param[in] pt The point to be project.
189  * \param[in] method The projection method to be used.
190  * \param[in] required_neighbors The minimum number of neighbors required.
191  * \note If required_neighbors is 0 then any number of neighbors is allowed.
192  * \note If required_neighbors is not satisfied it projects to the mls plane.
193  * \return The MLSProjectionResults for the input data.
194  */
195  inline MLSProjectionResults
196  projectPoint (const Eigen::Vector3d &pt, ProjectionMethod method, int required_neighbors = 0) const;
197 
198  /**
199  * \brief Project the query point used to generate the mls surface about using the specified method.
200  * \param[in] method The projection method to be used.
201  * \param[in] required_neighbors The minimum number of neighbors required.
202  * \note If required_neighbors is 0 then any number of neighbors is allowed.
203  * \note If required_neighbors is not satisfied it projects to the mls plane.
204  * \return The MLSProjectionResults for the input data.
205  */
206  inline MLSProjectionResults
207  projectQueryPoint (ProjectionMethod method, int required_neighbors = 0) const;
208 
209  /** \brief Smooth a given point and its neighborghood using Moving Least Squares.
210  * \param[in] index the index of the query point in the input cloud
211  * \param[in] nn_indices the set of nearest neighbors indices for pt
212  * \param[in] search_radius the search radius used to find nearest neighbors for pt
213  * \param[in] polynomial_order the order of the polynomial to fit to the nearest neighbors
214  * \param[in] weight_func defines the weight function for the polynomial fit
215  */
216  template <typename PointT> void
218  pcl::index_t index,
219  const pcl::Indices &nn_indices,
220  double search_radius,
221  int polynomial_order = 2,
222  std::function<double(const double)> weight_func = {});
223 
224  Eigen::Vector3d query_point; /**< \brief The query point about which the mls surface was generated */
225  Eigen::Vector3d mean; /**< \brief The mean point of all the neighbors. */
226  Eigen::Vector3d plane_normal; /**< \brief The normal of the local plane of the query point. */
227  Eigen::Vector3d u_axis; /**< \brief The axis corresponding to the u-coordinates of the local plane of the query point. */
228  Eigen::Vector3d v_axis; /**< \brief The axis corresponding to the v-coordinates of the local plane of the query point. */
229  Eigen::VectorXd c_vec; /**< \brief The polynomial coefficients Example: z = c_vec[0] + c_vec[1]*v + c_vec[2]*v^2 + c_vec[3]*u + c_vec[4]*u*v + c_vec[5]*u^2 */
230  int num_neighbors; /**< \brief The number of neighbors used to create the mls surface. */
231  float curvature; /**< \brief The curvature at the query point. */
232  int order; /**< \brief The order of the polynomial. If order > 1 then use polynomial fit */
233  bool valid; /**< \brief If True, the mls results data is valid, otherwise False. */
235  private:
236  /**
237  * \brief The default weight function used when fitting a polynomial surface
238  * \param sq_dist the squared distance from a point to origin of the mls frame
239  * \param sq_mls_radius the squraed mls search radius used
240  * \return The weight for a point at squared distance from the origin of the mls frame
241  */
242  inline
243  double computeMLSWeight (const double sq_dist, const double sq_mls_radius) { return (std::exp (-sq_dist / sq_mls_radius)); }
244 
245  };
246 
247  /** \brief MovingLeastSquares represent an implementation of the MLS (Moving Least Squares) algorithm
248  * for data smoothing and improved normal estimation. It also contains methods for upsampling the
249  * resulting cloud based on the parametric fit.
250  * Reference paper: "Computing and Rendering Point Set Surfaces" by Marc Alexa, Johannes Behr,
251  * Daniel Cohen-Or, Shachar Fleishman, David Levin and Claudio T. Silva
252  * www.sci.utah.edu/~shachar/Publications/crpss.pdf
253  * \note There is a parallelized version of the processing step, using the OpenMP standard.
254  * Compared to the standard version, an overhead is incurred in terms of runtime and memory usage.
255  * The upsampling methods DISTINCT_CLOUD and VOXEL_GRID_DILATION are not parallelized completely,
256  * i.e. parts of the algorithm run on a single thread only.
257  * \author Zoltan Csaba Marton, Radu B. Rusu, Alexandru E. Ichim, Suat Gedikli, Robert Huitl
258  * \ingroup surface
259  */
260  template <typename PointInT, typename PointOutT>
261  class MovingLeastSquares : public CloudSurfaceProcessing<PointInT, PointOutT>
262  {
263  public:
264  typedef shared_ptr<MovingLeastSquares<PointInT, PointOutT> > Ptr;
265  typedef shared_ptr<const MovingLeastSquares<PointInT, PointOutT> > ConstPtr;
266 
272 
274  using KdTreePtr = typename KdTree::Ptr;
277 
281 
285 
286  using SearchMethod = std::function<int (pcl::index_t, double, pcl::Indices &, std::vector<float> &)>;
287 
289  {
290  NONE, /**< \brief No upsampling will be done, only the input points will be projected
291  to their own MLS surfaces. */
292  DISTINCT_CLOUD, /**< \brief Project the points of the distinct cloud to the MLS surface. */
293  SAMPLE_LOCAL_PLANE, /**< \brief The local plane of each input point will be sampled in a circular fashion
294  using the \ref upsampling_radius_ and the \ref upsampling_step_ parameters. */
295  RANDOM_UNIFORM_DENSITY, /**< \brief The local plane of each input point will be sampled using an uniform random
296  distribution such that the density of points is constant throughout the
297  cloud - given by the \ref desired_num_points_in_radius_ parameter. */
298  VOXEL_GRID_DILATION /**< \brief The input cloud will be inserted into a voxel grid with voxels of
299  size \ref voxel_size_; this voxel grid will be dilated \ref dilation_iteration_num_
300  times and the resulting points will be projected to the MLS surface
301  of the closest point in the input cloud; the result is a point cloud
302  with filled holes and a constant point density. */
303  };
304 
305  /** \brief Empty constructor. */
306  MovingLeastSquares () : CloudSurfaceProcessing<PointInT, PointOutT> (),
307  distinct_cloud_ (),
308  tree_ (),
309  order_ (2),
310  search_radius_ (0.0),
311  sqr_gauss_param_ (0.0),
312  compute_normals_ (false),
314  upsampling_radius_ (0.0),
315  upsampling_step_ (0.0),
317  cache_mls_results_ (true),
318  projection_method_ (MLSResult::SIMPLE),
319  threads_ (1),
320  voxel_size_ (1.0),
322  nr_coeff_ (),
323  rng_uniform_distribution_ ()
324  {};
325 
326  /** \brief Empty destructor */
328 
329 
330  /** \brief Set whether the algorithm should also store the normals computed
331  * \note This is optional, but need a proper output cloud type
332  */
333  inline void
334  setComputeNormals (bool compute_normals) { compute_normals_ = compute_normals; }
335 
336  /** \brief Provide a pointer to the search object.
337  * \param[in] tree a pointer to the spatial search object.
338  */
339  inline void
341  {
342  tree_ = tree;
343  // Declare the search locator definition
344  search_method_ = [this] (pcl::index_t index, double radius, pcl::Indices& k_indices, std::vector<float>& k_sqr_distances)
345  {
346  return tree_->radiusSearch (index, radius, k_indices, k_sqr_distances, 0);
347  };
348  }
349 
350  /** \brief Get a pointer to the search method used. */
351  inline KdTreePtr
352  getSearchMethod () const { return (tree_); }
353 
354  /** \brief Set the order of the polynomial to be fit.
355  * \param[in] order the order of the polynomial
356  * \note Setting order > 1 indicates using a polynomial fit.
357  */
358  inline void
359  setPolynomialOrder (int order) { order_ = order; }
360 
361  /** \brief Get the order of the polynomial to be fit. */
362  inline int
363  getPolynomialOrder () const { return (order_); }
364 
365  /** \brief Set the sphere radius that is to be used for determining the k-nearest neighbors used for fitting.
366  * \param[in] radius the sphere radius that is to contain all k-nearest neighbors
367  * \note Calling this method resets the squared Gaussian parameter to radius * radius !
368  */
369  inline void
371 
372  /** \brief Get the sphere radius used for determining the k-nearest neighbors. */
373  inline double
374  getSearchRadius () const { return (search_radius_); }
375 
376  /** \brief Set the parameter used for distance based weighting of neighbors (the square of the search radius works
377  * best in general).
378  * \param[in] sqr_gauss_param the squared Gaussian parameter
379  */
380  inline void
381  setSqrGaussParam (double sqr_gauss_param) { sqr_gauss_param_ = sqr_gauss_param; }
382 
383  /** \brief Get the parameter for distance based weighting of neighbors. */
384  inline double
385  getSqrGaussParam () const { return (sqr_gauss_param_); }
386 
387  /** \brief Set the upsampling method to be used
388  * \param method
389  */
390  inline void
392 
393  /** \brief Set the distinct cloud used for the DISTINCT_CLOUD upsampling method. */
394  inline void
395  setDistinctCloud (PointCloudInConstPtr distinct_cloud) { distinct_cloud_ = distinct_cloud; }
396 
397  /** \brief Get the distinct cloud used for the DISTINCT_CLOUD upsampling method. */
398  inline PointCloudInConstPtr
399  getDistinctCloud () const { return (distinct_cloud_); }
400 
401 
402  /** \brief Set the radius of the circle in the local point plane that will be sampled
403  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
404  * \param[in] radius the radius of the circle
405  */
406  inline void
407  setUpsamplingRadius (double radius) { upsampling_radius_ = radius; }
408 
409  /** \brief Get the radius of the circle in the local point plane that will be sampled
410  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
411  */
412  inline double
414 
415  /** \brief Set the step size for the local plane sampling
416  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
417  * \param[in] step_size the step size
418  */
419  inline void
420  setUpsamplingStepSize (double step_size) { upsampling_step_ = step_size; }
421 
422 
423  /** \brief Get the step size for the local plane sampling
424  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
425  */
426  inline double
428 
429  /** \brief Set the parameter that specifies the desired number of points within the search radius
430  * \note Used only in the case of RANDOM_UNIFORM_DENSITY upsampling
431  * \param[in] desired_num_points_in_radius the desired number of points in the output cloud in a sphere of
432  * radius \ref search_radius_ around each point
433  */
434  inline void
435  setPointDensity (int desired_num_points_in_radius) { desired_num_points_in_radius_ = desired_num_points_in_radius; }
436 
437 
438  /** \brief Get the parameter that specifies the desired number of points within the search radius
439  * \note Used only in the case of RANDOM_UNIFORM_DENSITY upsampling
440  */
441  inline int
443 
444  /** \brief Set the voxel size for the voxel grid
445  * \note Used only in the VOXEL_GRID_DILATION upsampling method
446  * \param[in] voxel_size the edge length of a cubic voxel in the voxel grid
447  */
448  inline void
449  setDilationVoxelSize (float voxel_size) { voxel_size_ = voxel_size; }
450 
451 
452  /** \brief Get the voxel size for the voxel grid
453  * \note Used only in the VOXEL_GRID_DILATION upsampling method
454  */
455  inline float
456  getDilationVoxelSize () const { return (voxel_size_); }
457 
458  /** \brief Set the number of dilation steps of the voxel grid
459  * \note Used only in the VOXEL_GRID_DILATION upsampling method
460  * \param[in] iterations the number of dilation iterations
461  */
462  inline void
463  setDilationIterations (int iterations) { dilation_iteration_num_ = iterations; }
464 
465  /** \brief Get the number of dilation steps of the voxel grid
466  * \note Used only in the VOXEL_GRID_DILATION upsampling method
467  */
468  inline int
470 
471  /** \brief Set whether the mls results should be stored for each point in the input cloud
472  * \param[in] cache_mls_results True if the mls results should be stored, otherwise false.
473  * \note The cache_mls_results_ is forced to be true when using upsampling method VOXEL_GRID_DILATION or DISTINCT_CLOUD.
474  * \note If memory consumption is a concern, then set it to false when not using upsampling method VOXEL_GRID_DILATION or DISTINCT_CLOUD.
475  */
476  inline void
477  setCacheMLSResults (bool cache_mls_results) { cache_mls_results_ = cache_mls_results; }
478 
479  /** \brief Get the cache_mls_results_ value (True if the mls results should be stored, otherwise false). */
480  inline bool
481  getCacheMLSResults () const { return (cache_mls_results_); }
482 
483  /** \brief Set the method to be used when projection the point on to the MLS surface.
484  * \param method
485  * \note This is only used when polynomial fit is enabled.
486  */
487  inline void
489 
490 
491  /** \brief Get the current projection method being used. */
494 
495  /** \brief Get the MLSResults for input cloud
496  * \note The results are only stored if setCacheMLSResults(true) was called or when using the upsampling method DISTINCT_CLOUD or VOXEL_GRID_DILATION.
497  * \note This vector is aligned with the input cloud indices, so use getCorrespondingIndices to get the correct results when using output cloud indices.
498  */
499  inline const std::vector<MLSResult>&
500  getMLSResults () const { return (mls_results_); }
501 
502  /** \brief Set the maximum number of threads to use
503  * \param threads the maximum number of hardware threads to use (0 sets the value to 1)
504  */
505  inline void
506  setNumberOfThreads (unsigned int threads = 1)
507  {
508  threads_ = threads;
509  }
510 
511  /** \brief Base method for surface reconstruction for all points given in <setInputCloud (), setIndices ()>
512  * \param[out] output the resultant reconstructed surface model
513  */
514  void
515  process (PointCloudOut &output) override;
516 
517 
518  /** \brief Get the set of indices with each point in output having the
519  * corresponding point in input */
520  inline PointIndicesPtr
522 
523  protected:
524  /** \brief The point cloud that will hold the estimated normals, if set. */
526 
527  /** \brief The distinct point cloud that will be projected to the MLS surface. */
529 
530  /** \brief The search method template for indices. */
532 
533  /** \brief A pointer to the spatial search object. */
535 
536  /** \brief The order of the polynomial to be fit. */
537  int order_;
538 
539  /** \brief The nearest neighbors search radius for each point. */
541 
542  /** \brief Parameter for distance based weighting of neighbors (search_radius_ * search_radius_ works fine) */
544 
545  /** \brief Parameter that specifies whether the normals should be computed for the input cloud or not */
547 
548  /** \brief Parameter that specifies the upsampling method to be used */
550 
551  /** \brief Radius of the circle in the local point plane that will be sampled
552  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
553  */
555 
556  /** \brief Step size for the local plane sampling
557  * \note Used only in the case of SAMPLE_LOCAL_PLANE upsampling
558  */
560 
561  /** \brief Parameter that specifies the desired number of points within the search radius
562  * \note Used only in the case of RANDOM_UNIFORM_DENSITY upsampling
563  */
565 
566  /** \brief True if the mls results for the input cloud should be stored
567  * \note This is forced to be true when using upsampling methods VOXEL_GRID_DILATION or DISTINCT_CLOUD.
568  */
570 
571  /** \brief Stores the MLS result for each point in the input cloud
572  * \note Used only in the case of VOXEL_GRID_DILATION or DISTINCT_CLOUD upsampling
573  */
574  std::vector<MLSResult> mls_results_;
575 
576  /** \brief Parameter that specifies the projection method to be used. */
578 
579  /** \brief The maximum number of threads the scheduler should use. */
580  unsigned int threads_;
581 
582 
583  /** \brief A minimalistic implementation of a voxel grid, necessary for the point cloud upsampling
584  * \note Used only in the case of VOXEL_GRID_DILATION upsampling
585  */
587  {
588  public:
589  struct Leaf { Leaf () : valid (true) {} bool valid; };
590 
592  IndicesPtr &indices,
593  float voxel_size);
594 
595  void
596  dilate ();
597 
598  inline void
599  getIndexIn1D (const Eigen::Vector3i &index, std::uint64_t &index_1d) const
600  {
601  index_1d = index[0] * data_size_ * data_size_ +
602  index[1] * data_size_ + index[2];
603  }
604 
605  inline void
606  getIndexIn3D (std::uint64_t index_1d, Eigen::Vector3i& index_3d) const
607  {
608  index_3d[0] = static_cast<Eigen::Vector3i::Scalar> (index_1d / (data_size_ * data_size_));
609  index_1d -= index_3d[0] * data_size_ * data_size_;
610  index_3d[1] = static_cast<Eigen::Vector3i::Scalar> (index_1d / data_size_);
611  index_1d -= index_3d[1] * data_size_;
612  index_3d[2] = static_cast<Eigen::Vector3i::Scalar> (index_1d);
613  }
614 
615  inline void
616  getCellIndex (const Eigen::Vector3f &p, Eigen::Vector3i& index) const
617  {
618  for (int i = 0; i < 3; ++i)
619  index[i] = static_cast<Eigen::Vector3i::Scalar> ((p[i] - bounding_min_ (i)) / voxel_size_);
620  }
621 
622  inline void
623  getPosition (const std::uint64_t &index_1d, Eigen::Vector3f &point) const
624  {
625  Eigen::Vector3i index_3d;
626  getIndexIn3D (index_1d, index_3d);
627  for (int i = 0; i < 3; ++i)
628  point[i] = static_cast<Eigen::Vector3f::Scalar> (index_3d[i]) * voxel_size_ + bounding_min_[i];
629  }
630 
631  typedef std::map<std::uint64_t, Leaf> HashMap;
633  Eigen::Vector4f bounding_min_, bounding_max_;
634  std::uint64_t data_size_;
635  float voxel_size_;
637  };
638 
639 
640  /** \brief Voxel size for the VOXEL_GRID_DILATION upsampling method */
641  float voxel_size_;
642 
643  /** \brief Number of dilation steps for the VOXEL_GRID_DILATION upsampling method */
645 
646  /** \brief Number of coefficients, to be computed from the requested order.*/
648 
649  /** \brief Collects for each point in output the corrseponding point in the input. */
651 
652  /** \brief Search for the nearest neighbors of a given point using a radius search
653  * \param[in] index the index of the query point
654  * \param[out] indices the resultant vector of indices representing the neighbors within search_radius_
655  * \param[out] sqr_distances the resultant squared distances from the query point to the neighbors within search_radius_
656  */
657  inline int
658  searchForNeighbors (pcl::index_t index, pcl::Indices &indices, std::vector<float> &sqr_distances) const
659  {
660  return (search_method_ (index, search_radius_, indices, sqr_distances));
661  }
662 
663  /** \brief Smooth a given point and its neighborghood using Moving Least Squares.
664  * \param[in] index the index of the query point in the input cloud
665  * \param[in] nn_indices the set of nearest neighbors indices for pt
666  * \param[out] projected_points the set of projected points around the query point
667  * (in the case of upsampling method NONE, only the query point projected to its own fitted surface will be returned,
668  * in the case of the other upsampling methods, multiple points will be returned)
669  * \param[out] projected_points_normals the normals corresponding to the projected points
670  * \param[out] corresponding_input_indices the set of indices with each point in output having the corresponding point in input
671  * \param[out] mls_result stores the MLS result for each point in the input cloud
672  * (used only in the case of VOXEL_GRID_DILATION or DISTINCT_CLOUD upsampling)
673  */
674  void
676  const pcl::Indices &nn_indices,
677  PointCloudOut &projected_points,
678  NormalCloud &projected_points_normals,
679  PointIndices &corresponding_input_indices,
680  MLSResult &mls_result) const;
681 
682 
683  /** \brief This is a helper function for adding projected points
684  * \param[in] index the index of the query point in the input cloud
685  * \param[in] point the projected point to be added
686  * \param[in] normal the projected point's normal to be added
687  * \param[in] curvature the projected point's curvature
688  * \param[out] projected_points the set of projected points around the query point
689  * \param[out] projected_points_normals the normals corresponding to the projected points
690  * \param[out] corresponding_input_indices the set of indices with each point in output having the corresponding point in input
691  */
692  void
694  const Eigen::Vector3d &point,
695  const Eigen::Vector3d &normal,
696  double curvature,
697  PointCloudOut &projected_points,
698  NormalCloud &projected_points_normals,
699  PointIndices &corresponding_input_indices) const;
700 
701 
702  void
703  copyMissingFields (const PointInT &point_in,
704  PointOutT &point_out) const;
705 
706  /** \brief Abstract surface reconstruction method.
707  * \param[out] output the result of the reconstruction
708  */
709  void
710  performProcessing (PointCloudOut &output) override;
711 
712  /** \brief Perform upsampling for the distinct-cloud and voxel-grid methods
713  * \param[out] output the result of the reconstruction
714  */
715  void
717 
718  private:
719  /** \brief Random number generator algorithm. */
720  mutable std::mt19937 rng_;
721 
722  /** \brief Random number generator using an uniform distribution of floats
723  * \note Used only in the case of RANDOM_UNIFORM_DENSITY upsampling
724  */
725  std::unique_ptr<std::uniform_real_distribution<>> rng_uniform_distribution_;
726 
727  /** \brief Abstract class get name method. */
728  std::string
729  getClassName () const { return ("MovingLeastSquares"); }
730  };
731 }
732 
733 #ifdef PCL_NO_PRECOMPILE
734 #include <pcl/surface/impl/mls.hpp>
735 #endif
CloudSurfaceProcessing represents the base class for algorithms that takes a point cloud as input and...
Definition: processing.h:58
A minimalistic implementation of a voxel grid, necessary for the point cloud upsampling.
Definition: mls.h:587
MLSVoxelGrid(PointCloudInConstPtr &cloud, IndicesPtr &indices, float voxel_size)
Definition: mls.hpp:806
void getPosition(const std::uint64_t &index_1d, Eigen::Vector3f &point) const
Definition: mls.h:623
Eigen::Vector4f bounding_min_
Definition: mls.h:633
void getIndexIn1D(const Eigen::Vector3i &index, std::uint64_t &index_1d) const
Definition: mls.h:599
std::map< std::uint64_t, Leaf > HashMap
Definition: mls.h:631
Eigen::Vector4f bounding_max_
Definition: mls.h:633
void getCellIndex(const Eigen::Vector3f &p, Eigen::Vector3i &index) const
Definition: mls.h:616
void getIndexIn3D(std::uint64_t index_1d, Eigen::Vector3i &index_3d) const
Definition: mls.h:606
MovingLeastSquares represent an implementation of the MLS (Moving Least Squares) algorithm for data s...
Definition: mls.h:262
void setSqrGaussParam(double sqr_gauss_param)
Set the parameter used for distance based weighting of neighbors (the square of the search radius wor...
Definition: mls.h:381
void setDilationIterations(int iterations)
Set the number of dilation steps of the voxel grid.
Definition: mls.h:463
bool getCacheMLSResults() const
Get the cache_mls_results_ value (True if the mls results should be stored, otherwise false).
Definition: mls.h:481
double getSqrGaussParam() const
Get the parameter for distance based weighting of neighbors.
Definition: mls.h:385
unsigned int threads_
The maximum number of threads the scheduler should use.
Definition: mls.h:580
void performUpsampling(PointCloudOut &output)
Perform upsampling for the distinct-cloud and voxel-grid methods.
Definition: mls.hpp:370
~MovingLeastSquares()
Empty destructor.
Definition: mls.h:327
typename PointCloudIn::Ptr PointCloudInPtr
Definition: mls.h:283
int order_
The order of the polynomial to be fit.
Definition: mls.h:537
double getSearchRadius() const
Get the sphere radius used for determining the k-nearest neighbors.
Definition: mls.h:374
typename KdTree::Ptr KdTreePtr
Definition: mls.h:274
MLSResult::ProjectionMethod projection_method_
Parameter that specifies the projection method to be used.
Definition: mls.h:577
typename PointCloudOut::Ptr PointCloudOutPtr
Definition: mls.h:279
KdTreePtr getSearchMethod() const
Get a pointer to the search method used.
Definition: mls.h:352
void setPolynomialOrder(int order)
Set the order of the polynomial to be fit.
Definition: mls.h:359
int getPolynomialOrder() const
Get the order of the polynomial to be fit.
Definition: mls.h:363
double getUpsamplingRadius() const
Get the radius of the circle in the local point plane that will be sampled.
Definition: mls.h:413
double search_radius_
The nearest neighbors search radius for each point.
Definition: mls.h:540
MovingLeastSquares()
Empty constructor.
Definition: mls.h:306
pcl::PointCloud< PointOutT > PointCloudOut
Definition: mls.h:278
double sqr_gauss_param_
Parameter for distance based weighting of neighbors (search_radius_ * search_radius_ works fine)
Definition: mls.h:543
typename PointCloudIn::ConstPtr PointCloudInConstPtr
Definition: mls.h:284
int getPointDensity() const
Get the parameter that specifies the desired number of points within the search radius.
Definition: mls.h:442
std::function< int(pcl::index_t, double, pcl::Indices &, std::vector< float > &)> SearchMethod
Definition: mls.h:286
float getDilationVoxelSize() const
Get the voxel size for the voxel grid.
Definition: mls.h:456
KdTreePtr tree_
A pointer to the spatial search object.
Definition: mls.h:534
void copyMissingFields(const PointInT &point_in, PointOutT &point_out) const
Definition: mls.hpp:861
void setComputeNormals(bool compute_normals)
Set whether the algorithm should also store the normals computed.
Definition: mls.h:334
void setPointDensity(int desired_num_points_in_radius)
Set the parameter that specifies the desired number of points within the search radius.
Definition: mls.h:435
MLSResult::ProjectionMethod getProjectionMethod() const
Get the current projection method being used.
Definition: mls.h:493
double getUpsamplingStepSize() const
Get the step size for the local plane sampling.
Definition: mls.h:427
int getDilationIterations() const
Get the number of dilation steps of the voxel grid.
Definition: mls.h:469
double upsampling_step_
Step size for the local plane sampling.
Definition: mls.h:559
NormalCloud::Ptr NormalCloudPtr
Definition: mls.h:276
shared_ptr< MovingLeastSquares< PointInT, PointOutT > > Ptr
Definition: mls.h:264
void setUpsamplingRadius(double radius)
Set the radius of the circle in the local point plane that will be sampled.
Definition: mls.h:407
NormalCloudPtr normals_
The point cloud that will hold the estimated normals, if set.
Definition: mls.h:525
void setDistinctCloud(PointCloudInConstPtr distinct_cloud)
Set the distinct cloud used for the DISTINCT_CLOUD upsampling method.
Definition: mls.h:395
void setDilationVoxelSize(float voxel_size)
Set the voxel size for the voxel grid.
Definition: mls.h:449
UpsamplingMethod upsample_method_
Parameter that specifies the upsampling method to be used.
Definition: mls.h:549
int searchForNeighbors(pcl::index_t index, pcl::Indices &indices, std::vector< float > &sqr_distances) const
Search for the nearest neighbors of a given point using a radius search.
Definition: mls.h:658
double upsampling_radius_
Radius of the circle in the local point plane that will be sampled.
Definition: mls.h:554
@ RANDOM_UNIFORM_DENSITY
The local plane of each input point will be sampled using an uniform random distribution such that th...
Definition: mls.h:295
@ SAMPLE_LOCAL_PLANE
The local plane of each input point will be sampled in a circular fashion using the upsampling_radius...
Definition: mls.h:293
@ VOXEL_GRID_DILATION
The input cloud will be inserted into a voxel grid with voxels of size voxel_size_; this voxel grid w...
Definition: mls.h:298
@ NONE
No upsampling will be done, only the input points will be projected to their own MLS surfaces.
Definition: mls.h:290
@ DISTINCT_CLOUD
Project the points of the distinct cloud to the MLS surface.
Definition: mls.h:292
shared_ptr< const MovingLeastSquares< PointInT, PointOutT > > ConstPtr
Definition: mls.h:265
PointCloudInConstPtr getDistinctCloud() const
Get the distinct cloud used for the DISTINCT_CLOUD upsampling method.
Definition: mls.h:399
void setSearchMethod(const KdTreePtr &tree)
Provide a pointer to the search object.
Definition: mls.h:340
int desired_num_points_in_radius_
Parameter that specifies the desired number of points within the search radius.
Definition: mls.h:564
pcl::PointCloud< pcl::Normal > NormalCloud
Definition: mls.h:275
const std::vector< MLSResult > & getMLSResults() const
Get the MLSResults for input cloud.
Definition: mls.h:500
void setUpsamplingMethod(UpsamplingMethod method)
Set the upsampling method to be used.
Definition: mls.h:391
void setUpsamplingStepSize(double step_size)
Set the step size for the local plane sampling.
Definition: mls.h:420
PointIndicesPtr corresponding_input_indices_
Collects for each point in output the corrseponding point in the input.
Definition: mls.h:650
void setCacheMLSResults(bool cache_mls_results)
Set whether the mls results should be stored for each point in the input cloud.
Definition: mls.h:477
int nr_coeff_
Number of coefficients, to be computed from the requested order.
Definition: mls.h:647
void setNumberOfThreads(unsigned int threads=1)
Set the maximum number of threads to use.
Definition: mls.h:506
bool compute_normals_
Parameter that specifies whether the normals should be computed for the input cloud or not.
Definition: mls.h:546
void performProcessing(PointCloudOut &output) override
Abstract surface reconstruction method.
Definition: mls.hpp:284
void computeMLSPointNormal(pcl::index_t index, const pcl::Indices &nn_indices, PointCloudOut &projected_points, NormalCloud &projected_points_normals, PointIndices &corresponding_input_indices, MLSResult &mls_result) const
Smooth a given point and its neighborghood using Moving Least Squares.
Definition: mls.hpp:174
SearchMethod search_method_
The search method template for indices.
Definition: mls.h:531
void process(PointCloudOut &output) override
Base method for surface reconstruction for all points given in <setInputCloud (), setIndices ()>
Definition: mls.hpp:61
void addProjectedPointNormal(pcl::index_t index, const Eigen::Vector3d &point, const Eigen::Vector3d &normal, double curvature, PointCloudOut &projected_points, NormalCloud &projected_points_normals, PointIndices &corresponding_input_indices) const
This is a helper function for adding projected points.
Definition: mls.hpp:252
PointIndicesPtr getCorrespondingIndices() const
Get the set of indices with each point in output having the corresponding point in input.
Definition: mls.h:521
void setSearchRadius(double radius)
Set the sphere radius that is to be used for determining the k-nearest neighbors used for fitting.
Definition: mls.h:370
typename PointCloudOut::ConstPtr PointCloudOutConstPtr
Definition: mls.h:280
int dilation_iteration_num_
Number of dilation steps for the VOXEL_GRID_DILATION upsampling method.
Definition: mls.h:644
void setProjectionMethod(MLSResult::ProjectionMethod method)
Set the method to be used when projection the point on to the MLS surface.
Definition: mls.h:488
bool cache_mls_results_
True if the mls results for the input cloud should be stored.
Definition: mls.h:569
std::vector< MLSResult > mls_results_
Stores the MLS result for each point in the input cloud.
Definition: mls.h:574
float voxel_size_
Voxel size for the VOXEL_GRID_DILATION upsampling method.
Definition: mls.h:641
PointCloudInConstPtr distinct_cloud_
The distinct point cloud that will be projected to the MLS surface.
Definition: mls.h:528
PCL base class.
Definition: pcl_base.h:70
PointIndices::Ptr PointIndicesPtr
Definition: pcl_base.h:76
PointCloud represents the base class in PCL for storing collections of 3D points.
Definition: point_cloud.h:173
shared_ptr< PointCloud< pcl::Normal > > Ptr
Definition: point_cloud.h:413
shared_ptr< const PointCloud< PointOutT > > ConstPtr
Definition: point_cloud.h:414
shared_ptr< pcl::search::Search< PointInT > > Ptr
Definition: search.h:81
#define PCL_MAKE_ALIGNED_OPERATOR_NEW
Macro to signal a class requires a custom allocator.
Definition: memory.h:63
Defines functions, macros and traits for allocating and using memory.
Definition: bfgs.h:10
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
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition: types.h:133
PointIndices::Ptr PointIndicesPtr
Definition: PointIndices.h:24
shared_ptr< Indices > IndicesPtr
Definition: pcl_base.h:58
Defines all the PCL and non-PCL macros used.
#define PCL_DEPRECATED(Major, Minor, Message)
macro for compatibility across compilers and help remove old deprecated items for the Major....
Definition: pcl_macros.h:156
Data structure used to store the MLS projection results.
Definition: mls.h:81
Eigen::Vector3d point
The projected point.
Definition: mls.h:86
double v
The v-coordinate of the projected point in local MLS frame.
Definition: mls.h:85
Eigen::Vector3d normal
The projected point's normal.
Definition: mls.h:87
double u
The u-coordinate of the projected point in local MLS frame.
Definition: mls.h:84
Data structure used to store the MLS polynomial partial derivatives.
Definition: mls.h:70
double z_uv
The partial derivative d^2z/dudv.
Definition: mls.h:76
double z_u
The partial derivative dz/du.
Definition: mls.h:72
double z_uu
The partial derivative d^2z/du^2.
Definition: mls.h:74
double z
The z component of the polynomial evaluated at z(u, v).
Definition: mls.h:71
double z_vv
The partial derivative d^2z/dv^2.
Definition: mls.h:75
double z_v
The partial derivative dz/dv.
Definition: mls.h:73
Data structure used to store the results of the MLS fitting.
Definition: mls.h:60
MLSProjectionResults projectPoint(const Eigen::Vector3d &pt, ProjectionMethod method, int required_neighbors=0) const
Project a point using the specified method.
Definition: mls.hpp:637
MLSResult()
Definition: mls.h:92
Eigen::Vector3d mean
The mean point of all the neighbors.
Definition: mls.h:225
MLSProjectionResults projectPointOrthogonalToPolynomialSurface(const double u, const double v, const double w) const
Project a point orthogonal to the polynomial surface.
Definition: mls.hpp:537
Eigen::Vector3d u_axis
The axis corresponding to the u-coordinates of the local plane of the query point.
Definition: mls.h:227
Eigen::Vector2f calculatePrincipleCurvatures(const double u, const double v) const
Calculate the principal curvatures using the polynomial surface.
Definition: mls.h:155
Eigen::Vector3d plane_normal
The normal of the local plane of the query point.
Definition: mls.h:226
ProjectionMethod
Definition: mls.h:62
@ ORTHOGONAL
Project to the closest point on the polynonomial surface.
Definition: mls.h:65
@ SIMPLE
Project along the mls plane normal to the polynomial surface.
Definition: mls.h:64
@ NONE
Project to the mls plane.
Definition: mls.h:63
Eigen::Vector3d v_axis
The axis corresponding to the v-coordinates of the local plane of the query point.
Definition: mls.h:228
int num_neighbors
The number of neighbors used to create the mls surface.
Definition: mls.h:230
Eigen::VectorXd c_vec
The polynomial coefficients Example: z = c_vec[0] + c_vec[1]*v + c_vec[2]*v^2 + c_vec[3]*u + c_vec[4]...
Definition: mls.h:229
void computeMLSSurface(const pcl::PointCloud< PointT > &cloud, pcl::index_t index, const pcl::Indices &nn_indices, double search_radius, int polynomial_order=2, std::function< double(const double)> weight_func={})
Smooth a given point and its neighborghood using Moving Least Squares.
Definition: mls.hpp:690
void getMLSCoordinates(const Eigen::Vector3d &pt, double &u, double &v, double &w) const
Given a point calculate its 3D location in the MLS frame.
Definition: mls.hpp:453
float curvature
The curvature at the query point.
Definition: mls.h:231
PolynomialPartialDerivative getPolynomialPartialDerivative(const double u, const double v) const
Calculate the polynomial's first and second partial derivatives.
Definition: mls.hpp:492
MLSProjectionResults projectPointSimpleToPolynomialSurface(const double u, const double v) const
Project a point along the MLS plane normal to the polynomial surface.
Definition: mls.hpp:614
MLSProjectionResults projectPointToMLSPlane(const double u, const double v) const
Project a point onto the MLS plane.
Definition: mls.hpp:602
Eigen::Vector2f calculatePrincipalCurvatures(const double u, const double v) const
Calculate the principal curvatures using the polynomial surface.
double getPolynomialValue(const double u, const double v) const
Calculate the polynomial.
Definition: mls.hpp:470
Eigen::Vector3d query_point
The query point about which the mls surface was generated.
Definition: mls.h:224
MLSProjectionResults projectQueryPoint(ProjectionMethod method, int required_neighbors=0) const
Project the query point used to generate the mls surface about using the specified method.
Definition: mls.hpp:659
int order
The order of the polynomial.
Definition: mls.h:232
bool valid
If True, the mls results data is valid, otherwise False.
Definition: mls.h:233