Point Cloud Library (PCL)  1.14.0-dev
lmeds.hpp
1 /*
2  * Software License Agreement (BSD License)
3  *
4  * Point Cloud Library (PCL) - www.pointclouds.org
5  * Copyright (c) 2009, 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 #ifndef PCL_SAMPLE_CONSENSUS_IMPL_LMEDS_H_
42 #define PCL_SAMPLE_CONSENSUS_IMPL_LMEDS_H_
43 
44 #include <pcl/sample_consensus/lmeds.h>
45 #include <pcl/common/common.h> // for computeMedian
46 
47 //////////////////////////////////////////////////////////////////////////
48 template <typename PointT> bool
50 {
51  // Warn and exit if no threshold was set
52  if (threshold_ == std::numeric_limits<double>::max())
53  {
54  PCL_ERROR ("[pcl::LeastMedianSquares::computeModel] No threshold set!\n");
55  return (false);
56  }
57 
58  iterations_ = 0;
59  double d_best_penalty = std::numeric_limits<double>::max();
60 
61  Indices selection;
62  Eigen::VectorXf model_coefficients (sac_model_->getModelSize ());
63  std::vector<double> distances;
64 
65  unsigned skipped_count = 0;
66  // suppress infinite loops by just allowing 10 x maximum allowed iterations for invalid model parameters!
67  const unsigned max_skip = max_iterations_ * 10;
68 
69  // Iterate
70  while ((iterations_ < max_iterations_) && (skipped_count < max_skip))
71  {
72  // Get X samples which satisfy the model criteria
73  sac_model_->getSamples (iterations_, selection);
74 
75  if (selection.empty ())
76  {
77  PCL_ERROR ("[pcl::LeastMedianSquares::computeModel] No samples could be selected!\n");
78  break;
79  }
80 
81  // Search for inliers in the point cloud for the current plane model M
82  if (!sac_model_->computeModelCoefficients (selection, model_coefficients))
83  {
84  //iterations_++;
85  ++skipped_count;
86  PCL_DEBUG ("[pcl::LeastMedianSquares::computeModel] The function computeModelCoefficients failed, so continue with next iteration.\n");
87  continue;
88  }
89 
90  double d_cur_penalty;
91  // d_cur_penalty = sum (min (dist, threshold))
92 
93  // Iterate through the 3d points and calculate the distances from them to the model
94  sac_model_->getDistancesToModel (model_coefficients, distances);
95 
96  // No distances? The model must not respect the user given constraints
97  if (distances.empty ())
98  {
99  //iterations_++;
100  ++skipped_count;
101  continue;
102  }
103  // Move all NaNs in distances to the end
104  const auto new_end = (sac_model_->getInputCloud()->is_dense ? distances.end() : std::partition (distances.begin(), distances.end(), [](double d){return !std::isnan (d);}));
105  const auto nr_valid_dists = std::distance (distances.begin (), new_end);
106 
107  // d_cur_penalty = median (distances)
108  PCL_DEBUG ("[pcl::LeastMedianSquares::computeModel] There are %lu valid distances remaining after removing NaN values.\n", nr_valid_dists);
109  if (nr_valid_dists == 0)
110  {
111  //iterations_++;
112  ++skipped_count;
113  continue;
114  }
115  d_cur_penalty = pcl::computeMedian (distances.begin (), new_end, static_cast<double(*)(double)>(std::sqrt));
116 
117  // Better match ?
118  if (d_cur_penalty < d_best_penalty)
119  {
120  d_best_penalty = d_cur_penalty;
121 
122  // Save the current model/coefficients selection as being the best so far
123  model_ = selection;
124  model_coefficients_ = model_coefficients;
125  }
126 
127  ++iterations_;
128  if (debug_verbosity_level > 1)
129  {
130  PCL_DEBUG ("[pcl::LeastMedianSquares::computeModel] Trial %d out of %d. Best penalty is %f.\n", iterations_, max_iterations_, d_best_penalty);
131  }
132  }
133 
134  if (model_.empty ())
135  {
136  if (debug_verbosity_level > 0)
137  {
138  PCL_DEBUG ("[pcl::LeastMedianSquares::computeModel] Unable to find a solution!\n");
139  }
140  return (false);
141  }
142 
143  // Classify the data points into inliers and outliers
144  // Sigma = 1.4826 * (1 + 5 / (n-d)) * sqrt (M)
145  // @note: See "Robust Regression Methods for Computer Vision: A Review"
146  //double sigma = 1.4826 * (1 + 5 / (sac_model_->getIndices ()->size () - best_model.size ())) * sqrt (d_best_penalty);
147  //double threshold = 2.5 * sigma;
148 
149  // Iterate through the 3d points and calculate the distances from them to the model again
150  sac_model_->getDistancesToModel (model_coefficients_, distances);
151  // No distances? The model must not respect the user given constraints
152  if (distances.empty ())
153  {
154  PCL_ERROR ("[pcl::LeastMedianSquares::computeModel] The model found failed to verify against the given constraints!\n");
155  return (false);
156  }
157 
158  Indices &indices = *sac_model_->getIndices ();
159 
160  if (distances.size () != indices.size ())
161  {
162  PCL_ERROR ("[pcl::LeastMedianSquares::computeModel] Estimated distances (%lu) differs than the normal of indices (%lu).\n", distances.size (), indices.size ());
163  return (false);
164  }
165 
166  inliers_.resize (distances.size ());
167  // Get the inliers for the best model found
168  std::size_t n_inliers_count = 0;
169  for (std::size_t i = 0; i < distances.size (); ++i)
170  {
171  if (distances[i] <= threshold_)
172  {
173  inliers_[n_inliers_count++] = indices[i];
174  }
175  }
176 
177  // Resize the inliers vector
178  inliers_.resize (n_inliers_count);
179 
180  if (debug_verbosity_level > 0)
181  {
182  PCL_DEBUG ("[pcl::LeastMedianSquares::computeModel] Model: %lu size, %lu inliers.\n", model_.size (), n_inliers_count);
183  }
184 
185  return (true);
186 }
187 
188 #define PCL_INSTANTIATE_LeastMedianSquares(T) template class PCL_EXPORTS pcl::LeastMedianSquares<T>;
189 
190 #endif // PCL_SAMPLE_CONSENSUS_IMPL_LMEDS_H_
bool computeModel(int debug_verbosity_level=0) override
Compute the actual model and find the inliers.
Definition: lmeds.hpp:49
Define standard C methods and C++ classes that are common to all methods.
auto computeMedian(IteratorT begin, IteratorT end, Functor f) noexcept -> std::result_of_t< Functor(decltype(*begin))>
Compute the median of a list of values (fast).
Definition: common.h:285
float distance(const PointT &p1, const PointT &p2)
Definition: geometry.h:60
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition: types.h:133