Point Cloud Library (PCL) 1.14.0
Loading...
Searching...
No Matches
correspondence_rejection_poly.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 */
38
39#pragma once
40
41#include <pcl/registration/correspondence_rejection.h>
42#include <pcl/point_cloud.h>
43
44namespace pcl {
45namespace registration {
46/** \brief CorrespondenceRejectorPoly implements a correspondence rejection method that
47 * exploits low-level and pose-invariant geometric constraints between two point sets by
48 * forming virtual polygons of a user-specifiable cardinality on each model using the
49 * input correspondences. These polygons are then checked in a pose-invariant manner
50 * (i.e. the side lengths must be approximately equal), and rejection is performed by
51 * thresholding these edge lengths.
52 *
53 * If you use this in academic work, please cite:
54 *
55 * A. G. Buch, D. Kraft, J.-K. Kämäräinen, H. G. Petersen and N. Krüger.
56 * Pose Estimation using Local Structure-Specific Shape and Appearance Context.
57 * International Conference on Robotics and Automation (ICRA), 2013.
58 *
59 * \author Anders Glent Buch
60 * \ingroup registration
61 */
62template <typename SourceT, typename TargetT>
64 using CorrespondenceRejector::getClassName;
65 using CorrespondenceRejector::input_correspondences_;
66 using CorrespondenceRejector::rejection_name_;
67
68public:
69 using Ptr = shared_ptr<CorrespondenceRejectorPoly<SourceT, TargetT>>;
70 using ConstPtr = shared_ptr<const CorrespondenceRejectorPoly<SourceT, TargetT>>;
71
73 using PointCloudSourcePtr = typename PointCloudSource::Ptr;
74 using PointCloudSourceConstPtr = typename PointCloudSource::ConstPtr;
75
77 using PointCloudTargetPtr = typename PointCloudTarget::Ptr;
78 using PointCloudTargetConstPtr = typename PointCloudTarget::ConstPtr;
79
80 /** \brief Empty constructor */
81 CorrespondenceRejectorPoly() { rejection_name_ = "CorrespondenceRejectorPoly"; }
82
83 /** \brief Get a list of valid correspondences after rejection from the original set
84 * of correspondences.
85 * \param[in] original_correspondences the set of initial correspondences given
86 * \param[out] remaining_correspondences the resultant filtered set of remaining
87 * correspondences
88 */
89 void
90 getRemainingCorrespondences(const pcl::Correspondences& original_correspondences,
91 pcl::Correspondences& remaining_correspondences) override;
92
93 /** \brief Provide a source point cloud dataset (must contain XYZ data!), used to
94 * compute the correspondence distance.
95 * \param[in] cloud a cloud containing XYZ data
96 */
97 inline void
99 {
100 input_ = cloud;
101 }
102
103 /** \brief Provide a target point cloud dataset (must contain XYZ data!), used to
104 * compute the correspondence distance.
105 * \param[in] target a cloud containing XYZ data
106 */
107 inline void
109 {
110 target_ = target;
111 }
112
113 /** \brief See if this rejector requires source points */
114 bool
115 requiresSourcePoints() const override
116 {
117 return (true);
118 }
119
120 /** \brief Blob method for setting the source cloud */
121 void
123 {
125 fromPCLPointCloud2(*cloud2, *cloud);
126 setInputSource(cloud);
127 }
128
129 /** \brief See if this rejector requires a target cloud */
130 bool
131 requiresTargetPoints() const override
132 {
133 return (true);
134 }
135
136 /** \brief Method for setting the target cloud */
137 void
139 {
141 fromPCLPointCloud2(*cloud2, *cloud);
142 setInputTarget(cloud);
143 }
144
145 /** \brief Set the polygon cardinality
146 * \param cardinality polygon cardinality
147 */
148 inline void
149 setCardinality(int cardinality)
150 {
151 cardinality_ = cardinality;
152 }
153
154 /** \brief Get the polygon cardinality
155 * \return polygon cardinality
156 */
157 inline int
159 {
160 return (cardinality_);
161 }
162
163 /** \brief Set the similarity threshold in [0,1[ between edge lengths,
164 * where 1 is a perfect match
165 * \param similarity_threshold similarity threshold
166 */
167 inline void
168 setSimilarityThreshold(float similarity_threshold)
169 {
170 similarity_threshold_ = similarity_threshold;
171 similarity_threshold_squared_ = similarity_threshold * similarity_threshold;
172 }
173
174 /** \brief Get the similarity threshold between edge lengths
175 * \return similarity threshold
176 */
177 inline float
179 {
180 return (similarity_threshold_);
181 }
182
183 /** \brief Set the number of iterations
184 * \param iterations number of iterations
185 */
186 inline void
187 setIterations(int iterations)
188 {
189 iterations_ = iterations;
190 }
191
192 /** \brief Get the number of iterations
193 * \return number of iterations
194 */
195 inline int
197 {
198 return (iterations_);
199 }
200
201 /** \brief Polygonal rejection of a single polygon, indexed by a subset of
202 * correspondences \param corr all correspondences into \ref input_ and \ref target_
203 * \param idx sampled indices into \b correspondences, must have a size equal to \ref
204 * cardinality_ \return true if all edge length ratios are larger than or equal to
205 * \ref similarity_threshold_
206 */
207 inline bool
208 thresholdPolygon(const pcl::Correspondences& corr, const std::vector<int>& idx)
209 {
210 if (cardinality_ ==
211 2) // Special case: when two points are considered, we only have one edge
212 {
213 return (thresholdEdgeLength(corr[idx[0]].index_query,
214 corr[idx[1]].index_query,
215 corr[idx[0]].index_match,
216 corr[idx[1]].index_match,
217 similarity_threshold_squared_));
218 }
219 // Otherwise check all edges
220 for (int i = 0; i < cardinality_; ++i) {
221 if (!thresholdEdgeLength(corr[idx[i]].index_query,
222 corr[idx[(i + 1) % cardinality_]].index_query,
223 corr[idx[i]].index_match,
224 corr[idx[(i + 1) % cardinality_]].index_match,
225 similarity_threshold_squared_)) {
226 return (false);
227 }
228 }
229 return (true);
230 }
231
232 /** \brief Polygonal rejection of a single polygon, indexed by two point index vectors
233 * \param source_indices indices of polygon points in \ref input_, must have a size
234 * equal to \ref cardinality_
235 * \param target_indices corresponding indices of polygon points in \ref target_, must
236 * have a size equal to \ref cardinality_
237 * \return true if all edge length ratios are larger than or equal to
238 * \ref similarity_threshold_
239 */
240 inline bool
241 thresholdPolygon(const pcl::Indices& source_indices,
242 const pcl::Indices& target_indices)
243 {
244 // Convert indices to correspondences and an index vector pointing to each element
245 pcl::Correspondences corr(cardinality_);
246 std::vector<int> idx(cardinality_);
247 for (int i = 0; i < cardinality_; ++i) {
248 corr[i].index_query = source_indices[i];
249 corr[i].index_match = target_indices[i];
250 idx[i] = i;
251 }
252
253 return (thresholdPolygon(corr, idx));
254 }
255
256protected:
257 /** \brief Apply the rejection algorithm.
258 * \param[out] correspondences the set of resultant correspondences.
259 */
260 inline void
261 applyRejection(pcl::Correspondences& correspondences) override
262 {
263 getRemainingCorrespondences(*input_correspondences_, correspondences);
264 }
265
266 /** \brief Get k unique random indices in range {0,...,n-1} (sampling without
267 * replacement) \note No check is made to ensure that k <= n.
268 * \param n upper index range, exclusive
269 * \param k number of unique indices to sample
270 * \return k unique random indices in range {0,...,n-1}
271 */
272 inline std::vector<int>
274 {
275 // Marked sampled indices and sample counter
276 std::vector<bool> sampled(n, false);
277 int samples = 0;
278 // Resulting unique indices
279 std::vector<int> result;
280 result.reserve(k);
281 do {
282 // Pick a random index in the range
283 const int idx = (std::rand() % n);
284 // If unique
285 if (!sampled[idx]) {
286 // Mark as sampled and increment result counter
287 sampled[idx] = true;
288 ++samples;
289 // Store
290 result.push_back(idx);
291 }
292 } while (samples < k);
293
294 return (result);
295 }
296
297 /** \brief Squared Euclidean distance between two points using the members x, y and z
298 * \param p1 first point
299 * \param p2 second point
300 * \return squared Euclidean distance
301 */
302 inline float
303 computeSquaredDistance(const SourceT& p1, const TargetT& p2)
304 {
305 const float dx = p2.x - p1.x;
306 const float dy = p2.y - p1.y;
307 const float dz = p2.z - p1.z;
308
309 return (dx * dx + dy * dy + dz * dz);
310 }
311
312 /** \brief Edge length similarity thresholding
313 * \param index_query_1 index of first source vertex
314 * \param index_query_2 index of second source vertex
315 * \param index_match_1 index of first target vertex
316 * \param index_match_2 index of second target vertex
317 * \param simsq squared similarity threshold in [0,1]
318 * \return true if edge length ratio is larger than or equal to threshold
319 */
320 inline bool
321 thresholdEdgeLength(int index_query_1,
322 int index_query_2,
323 int index_match_1,
324 int index_match_2,
325 float simsq)
326 {
327 // Distance between source points
328 const float dist_src =
329 computeSquaredDistance((*input_)[index_query_1], (*input_)[index_query_2]);
330 // Distance between target points
331 const float dist_tgt =
332 computeSquaredDistance((*target_)[index_match_1], (*target_)[index_match_2]);
333 // Edge length similarity [0,1] where 1 is a perfect match
334 const float edge_sim =
335 (dist_src < dist_tgt ? dist_src / dist_tgt : dist_tgt / dist_src);
336
337 return (edge_sim >= simsq);
338 }
339
340 /** \brief Compute a linear histogram. This function is equivalent to the MATLAB
341 * function \b histc, with the edges set as follows: <b>
342 * lower:(upper-lower)/bins:upper </b>
343 * \param data input samples
344 * \param lower lower bound of input samples
345 * \param upper upper bound of input samples
346 * \param bins number of bins in output
347 * \return linear histogram
348 */
349 std::vector<int>
350 computeHistogram(const std::vector<float>& data, float lower, float upper, int bins);
351
352 /** \brief Find the optimal value for binary histogram thresholding using Otsu's
353 * method
354 * \param histogram input histogram \return threshold value according to Otsu's
355 * criterion
356 */
357 int
358 findThresholdOtsu(const std::vector<int>& histogram);
359
360 /** \brief The input point cloud dataset */
362
363 /** \brief The input point cloud dataset target */
365
366 /** \brief Number of iterations to run */
367 int iterations_{10000};
368
369 /** \brief The polygon cardinality used during rejection */
370 int cardinality_{3};
371
372 /** \brief Lower edge length threshold in [0,1] used for verifying polygon
373 * similarities, where 1 is a perfect match */
374 float similarity_threshold_{0.75f};
375
376 /** \brief Squared value if \ref similarity_threshold_, only for internal use */
377 float similarity_threshold_squared_{0.75f * 0.75f};
378};
379} // namespace registration
380} // namespace pcl
381
382#include <pcl/registration/impl/correspondence_rejection_poly.hpp>
PointCloud represents the base class in PCL for storing collections of 3D points.
CorrespondenceRejector represents the base class for correspondence rejection methods
CorrespondenceRejectorPoly implements a correspondence rejection method that exploits low-level and p...
std::vector< int > getUniqueRandomIndices(int n, int k)
Get k unique random indices in range {0,...,n-1} (sampling without replacement)
bool thresholdPolygon(const pcl::Correspondences &corr, const std::vector< int > &idx)
Polygonal rejection of a single polygon, indexed by a subset of correspondences.
shared_ptr< const CorrespondenceRejectorPoly< SourceT, TargetT > > ConstPtr
float computeSquaredDistance(const SourceT &p1, const TargetT &p2)
Squared Euclidean distance between two points using the members x, y and z.
shared_ptr< CorrespondenceRejectorPoly< SourceT, TargetT > > Ptr
typename PointCloudTarget::ConstPtr PointCloudTargetConstPtr
PointCloudSourceConstPtr input_
The input point cloud dataset.
float getSimilarityThreshold()
Get the similarity threshold between edge lengths.
void setTargetPoints(pcl::PCLPointCloud2::ConstPtr cloud2) override
Method for setting the target cloud.
bool thresholdPolygon(const pcl::Indices &source_indices, const pcl::Indices &target_indices)
Polygonal rejection of a single polygon, indexed by two point index vectors.
void applyRejection(pcl::Correspondences &correspondences) override
Apply the rejection algorithm.
void setCardinality(int cardinality)
Set the polygon cardinality.
void setSimilarityThreshold(float similarity_threshold)
Set the similarity threshold in [0,1[ between edge lengths, where 1 is a perfect match.
void setIterations(int iterations)
Set the number of iterations.
bool thresholdEdgeLength(int index_query_1, int index_query_2, int index_match_1, int index_match_2, float simsq)
Edge length similarity thresholding.
PointCloudTargetConstPtr target_
The input point cloud dataset target.
void setInputTarget(const PointCloudTargetConstPtr &target)
Provide a target point cloud dataset (must contain XYZ data!), used to compute the correspondence dis...
void setSourcePoints(pcl::PCLPointCloud2::ConstPtr cloud2) override
Blob method for setting the source cloud.
void setInputSource(const PointCloudSourceConstPtr &cloud)
Provide a source point cloud dataset (must contain XYZ data!), used to compute the correspondence dis...
bool requiresTargetPoints() const override
See if this rejector requires a target cloud.
bool requiresSourcePoints() const override
See if this rejector requires source points.
typename PointCloudSource::ConstPtr PointCloudSourceConstPtr
void fromPCLPointCloud2(const pcl::PCLPointCloud2 &msg, pcl::PointCloud< PointT > &cloud, const MsgFieldMap &field_map, const std::uint8_t *msg_data)
Convert a PCLPointCloud2 binary data blob into a pcl::PointCloud<T> object using a field_map.
std::vector< pcl::Correspondence, Eigen::aligned_allocator< pcl::Correspondence > > Correspondences
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition types.h:133
shared_ptr< const ::pcl::PCLPointCloud2 > ConstPtr