Point Cloud Library (PCL) 1.12.0
Loading...
Searching...
No Matches
flare.hpp
1/*
2* Software License Agreement (BSD License)
3*
4* Point Cloud Library (PCL) - www.pointclouds.org
5* Copyright (c) 2016-, Open Perception, 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 the copyright holder(s) 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*
37*/
38
39#ifndef PCL_FEATURES_IMPL_FLARE_H_
40#define PCL_FEATURES_IMPL_FLARE_H_
41
42#include <pcl/features/flare.h>
43#include <pcl/common/geometry.h>
44
45//////////////////////////////////////////////////////////////////////////////////////////////
46template<typename PointInT, typename PointNT, typename PointOutT, typename SignedDistanceT> bool
48{
50 {
51 PCL_ERROR ("[pcl::%s::initCompute] Init failed.\n", getClassName ().c_str ());
52 return (false);
53 }
54
55 if (tangent_radius_ == 0.0f)
56 {
57 PCL_ERROR ("[pcl::%s::initCompute] tangent_radius_ not set.\n", getClassName ().c_str ());
58 return (false);
59 }
60
61 // If no search sampled_surface_ has been defined, use the surface_ dataset as the search sampled_surface_ itself
62 if (!sampled_surface_)
63 {
64 fake_sampled_surface_ = true;
65 sampled_surface_ = surface_;
66
67 if (sampled_tree_)
68 {
69 PCL_WARN ("[pcl::%s::initCompute] sampled_surface_ is not set even if sampled_tree_ is already set.", getClassName ().c_str ());
70 PCL_WARN ("sampled_tree_ will be rebuilt from surface_. Use sampled_surface_.\n");
71 }
72 }
73
74 // Check if a space search locator was given for sampled_surface_
75 if (!sampled_tree_)
76 {
77 if (sampled_surface_->isOrganized () && surface_->isOrganized () && input_->isOrganized ())
78 sampled_tree_.reset (new pcl::search::OrganizedNeighbor<PointInT> ());
79 else
80 sampled_tree_.reset (new pcl::search::KdTree<PointInT> (false));
81 }
82
83 if (sampled_tree_->getInputCloud () != sampled_surface_) // Make sure the tree searches the sampled surface
84 sampled_tree_->setInputCloud (sampled_surface_);
85
86 return (true);
87}
88
89//////////////////////////////////////////////////////////////////////////////////////////////
90template<typename PointInT, typename PointNT, typename PointOutT, typename SignedDistanceT> bool
92{
93 // Reset the surface
94 if (fake_surface_)
95 {
96 surface_.reset ();
97 fake_surface_ = false;
98 }
99 // Reset the sampled surface
100 if (fake_sampled_surface_)
101 {
102 sampled_surface_.reset ();
103 fake_sampled_surface_ = false;
104 }
105 return (true);
106}
107
108//////////////////////////////////////////////////////////////////////////////////////////////
109template<typename PointInT, typename PointNT, typename PointOutT, typename SignedDistanceT> SignedDistanceT
111 Eigen::Matrix3f &lrf)
112{
113 Eigen::Vector3f x_axis, y_axis;
114 Eigen::Vector3f fitted_normal; //z_axis
115
116 //find Z axis
117
118 //extract support points for the computation of Z axis
120 std::vector<float> neighbours_distances;
121
122 const std::size_t n_normal_neighbours =
123 this->searchForNeighbors (index, search_parameter_, neighbours_indices, neighbours_distances);
124 if (n_normal_neighbours < static_cast<std::size_t>(min_neighbors_for_normal_axis_))
125 {
126 if (!pcl::isFinite ((*normals_)[index]))
127 {
128 //normal is invalid
129 //setting lrf to NaN
130 lrf.setConstant (std::numeric_limits<float>::quiet_NaN ());
131 return (std::numeric_limits<SignedDistanceT>::max ());
132 }
133
134 //set z_axis as the normal of index point
135 fitted_normal = (*normals_)[index].getNormalVector3fMap ();
136 }
137 else
138 {
139 float plane_curvature;
140 normal_estimation_.computePointNormal (*surface_, neighbours_indices, fitted_normal (0), fitted_normal (1), fitted_normal (2), plane_curvature);
141
142 //disambiguate Z axis with normal mean
144 {
145 //all normals in the neighbourood are invalid
146 //setting lrf to NaN
147 lrf.setConstant (std::numeric_limits<float>::quiet_NaN ());
148 return (std::numeric_limits<SignedDistanceT>::max ());
149 }
150 }
151
152 //setting LRF Z axis
153 lrf.row (2).matrix () = fitted_normal;
154
155 //find X axis
156
157 //extract support points for Rx radius
158 const std::size_t n_tangent_neighbours =
159 sampled_tree_->radiusSearch ((*input_)[index], tangent_radius_, neighbours_indices, neighbours_distances);
160
161 if (n_tangent_neighbours < static_cast<std::size_t>(min_neighbors_for_tangent_axis_))
162 {
163 //set X axis as a random axis
165 y_axis = fitted_normal.cross (x_axis);
166
167 lrf.row (0).matrix () = x_axis;
168 lrf.row (1).matrix () = y_axis;
169
170 return (std::numeric_limits<SignedDistanceT>::max ());
171 }
172
173 //find point with the largest signed distance from tangent plane
174
176 SignedDistanceT best_shape_score = -std::numeric_limits<SignedDistanceT>::max ();
177 int best_shape_index = -1;
178
179 Eigen::Vector3f best_margin_point;
180
181 const float radius2 = tangent_radius_ * tangent_radius_;
182 const float margin_distance2 = margin_thresh_ * margin_thresh_ * radius2;
183
184 Vector3fMapConst feature_point = (*input_)[index].getVector3fMap ();
185
186 for (std::size_t curr_neigh = 0; curr_neigh < n_tangent_neighbours; ++curr_neigh)
187 {
190
192 {
193 continue;
194 }
195
196 //point curr_neigh_idx is inside the ring between marginThresh and Radius
197
198 shape_score = fitted_normal.dot ((*sampled_surface_)[curr_neigh_idx].getVector3fMap ());
199
201 {
204 }
205 } //for each neighbor
206
207 if (best_shape_index == -1)
208 {
210 y_axis = fitted_normal.cross (x_axis);
211
212 lrf.row (0).matrix () = x_axis;
213 lrf.row (1).matrix () = y_axis;
214
215 return (std::numeric_limits<SignedDistanceT>::max ());
216 }
217
218 //find orthogonal axis directed to best_shape_index point projection on plane with fittedNormal as axis
219 x_axis = pcl::geometry::projectedAsUnitVector (sampled_surface_->at (best_shape_index).getVector3fMap (), feature_point, fitted_normal);
220
221 y_axis = fitted_normal.cross (x_axis);
222
223 lrf.row (0).matrix () = x_axis;
224 lrf.row (1).matrix () = y_axis;
225 //z axis already set
226
228 return (best_shape_score);
229}
230
231//////////////////////////////////////////////////////////////////////////////////////////////
232template<typename PointInT, typename PointNT, typename PointOutT, typename SignedDistanceT> void
234{
235 //check whether used with search radius or search k-neighbors
236 if (this->getKSearch () != 0)
237 {
238 PCL_ERROR (
239 "[pcl::%s::computeFeature] Error! Search method set to k-neighborhood. Call setKSearch (0) and setRadiusSearch (radius) to use this class.\n",
240 getClassName ().c_str ());
241 return;
242 }
243
244 signed_distances_from_highest_points_.resize (indices_->size ());
245
246 for (std::size_t point_idx = 0; point_idx < indices_->size (); ++point_idx)
247 {
248 Eigen::Matrix3f currentLrf;
250
251 signed_distances_from_highest_points_[point_idx] = computePointLRF ((*indices_)[point_idx], currentLrf);
252 if (signed_distances_from_highest_points_[point_idx] == std::numeric_limits<SignedDistanceT>::max ())
253 {
254 output.is_dense = false;
255 }
256
257 rf.getXAxisVector3fMap () = currentLrf.row (0);
258 rf.getYAxisVector3fMap () = currentLrf.row (1);
259 rf.getZAxisVector3fMap () = currentLrf.row (2);
260 }
261}
262
263#define PCL_INSTANTIATE_FLARELocalReferenceFrameEstimation(T,NT,OutT,SdT) template class PCL_EXPORTS pcl::FLARELocalReferenceFrameEstimation<T,NT,OutT,SdT>;
264
265#endif // PCL_FEATURES_IMPL_FLARE_H_
Iterator class for point clouds with or without given indices.
std::size_t size() const
Size of the range the iterator is going through.
bool deinitCompute() override
This method should get called after the actual computation is ended.
Definition flare.hpp:91
void computeFeature(PointCloudOut &output) override
Abstract feature estimation method.
Definition flare.hpp:233
bool initCompute() override
This method should get called before starting the actual computation.
Definition flare.hpp:47
SignedDistanceT computePointLRF(const int index, Eigen::Matrix3f &lrf)
Estimate the LRF descriptor for a given point based on its spatial neighborhood of 3D points with nor...
Definition flare.hpp:110
search::KdTree is a wrapper class which inherits the pcl::KdTree class for performing search function...
Definition kdtree.h:62
OrganizedNeighbor is a class for optimized nearest neigbhor search in organized point clouds.
Definition organized.h:61
Defines some geometrical functions and utility functions.
Eigen::Vector3f projectedAsUnitVector(Eigen::Vector3f const &point, Eigen::Vector3f const &plane_origin, Eigen::Vector3f const &plane_normal)
Given a plane defined by plane_origin and plane_normal, find the unit vector pointing from plane_orig...
Definition geometry.h:115
Eigen::Vector3f randomOrthogonalAxis(Eigen::Vector3f const &axis)
Define a random unit vector orthogonal to axis.
Definition geometry.h:134
const Eigen::Map< const Eigen::Vector3f > Vector3fMapConst
bool isFinite(const PointT &pt)
Tests if the 3D components of a point are all finite param[in] pt point to be tested return true if f...
Definition point_tests.h:55