Point Cloud Library (PCL) 1.12.0
Loading...
Searching...
No Matches
transformation_estimation_svd.hpp
1/*
2 * Software License Agreement (BSD License)
3 *
4 * Point Cloud Library (PCL) - www.pointclouds.org
5 * Copyright (c) 2010, 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_REGISTRATION_TRANSFORMATION_ESTIMATION_SVD_HPP_
42#define PCL_REGISTRATION_TRANSFORMATION_ESTIMATION_SVD_HPP_
43
44#include <pcl/common/eigen.h>
45
46namespace pcl {
47
48namespace registration {
49
50template <typename PointSource, typename PointTarget, typename Scalar>
51inline void
54 const pcl::PointCloud<PointTarget>& cloud_tgt,
55 Matrix4& transformation_matrix) const
56{
57 const auto nr_points = cloud_src.size();
58 if (cloud_tgt.size() != nr_points) {
59 PCL_ERROR("[pcl::TransformationEstimationSVD::estimateRigidTransformation] Number "
60 "or points in source (%zu) differs than target (%zu)!\n",
61 static_cast<std::size_t>(nr_points),
62 static_cast<std::size_t>(cloud_tgt.size()));
63 return;
64 }
65
66 ConstCloudIterator<PointSource> source_it(cloud_src);
67 ConstCloudIterator<PointTarget> target_it(cloud_tgt);
68 estimateRigidTransformation(source_it, target_it, transformation_matrix);
69}
70
71template <typename PointSource, typename PointTarget, typename Scalar>
72void
75 const pcl::Indices& indices_src,
76 const pcl::PointCloud<PointTarget>& cloud_tgt,
77 Matrix4& transformation_matrix) const
78{
79 if (indices_src.size() != cloud_tgt.size()) {
80 PCL_ERROR("[pcl::TransformationSVD::estimateRigidTransformation] Number or points "
81 "in source (%zu) differs than target (%zu)!\n",
82 indices_src.size(),
83 static_cast<std::size_t>(cloud_tgt.size()));
84 return;
85 }
86
87 ConstCloudIterator<PointSource> source_it(cloud_src, indices_src);
88 ConstCloudIterator<PointTarget> target_it(cloud_tgt);
89 estimateRigidTransformation(source_it, target_it, transformation_matrix);
90}
92template <typename PointSource, typename PointTarget, typename Scalar>
93inline void
96 const pcl::Indices& indices_src,
97 const pcl::PointCloud<PointTarget>& cloud_tgt,
98 const pcl::Indices& indices_tgt,
99 Matrix4& transformation_matrix) const
100{
101 if (indices_src.size() != indices_tgt.size()) {
102 PCL_ERROR("[pcl::TransformationEstimationSVD::estimateRigidTransformation] Number "
103 "or points in source (%zu) differs than target (%zu)!\n",
104 indices_src.size(),
105 indices_tgt.size());
106 return;
107 }
108
109 ConstCloudIterator<PointSource> source_it(cloud_src, indices_src);
110 ConstCloudIterator<PointTarget> target_it(cloud_tgt, indices_tgt);
111 estimateRigidTransformation(source_it, target_it, transformation_matrix);
112}
113
114template <typename PointSource, typename PointTarget, typename Scalar>
115void
118 const pcl::PointCloud<PointTarget>& cloud_tgt,
119 const pcl::Correspondences& correspondences,
120 Matrix4& transformation_matrix) const
121{
122 ConstCloudIterator<PointSource> source_it(cloud_src, correspondences, true);
123 ConstCloudIterator<PointTarget> target_it(cloud_tgt, correspondences, false);
124 estimateRigidTransformation(source_it, target_it, transformation_matrix);
125}
126
127template <typename PointSource, typename PointTarget, typename Scalar>
128inline void
132 Matrix4& transformation_matrix) const
133{
134 // Convert to Eigen format
135 const int npts = static_cast<int>(source_it.size());
136
137 if (use_umeyama_) {
138 Eigen::Matrix<Scalar, 3, Eigen::Dynamic> cloud_src(3, npts);
139 Eigen::Matrix<Scalar, 3, Eigen::Dynamic> cloud_tgt(3, npts);
140
141 for (int i = 0; i < npts; ++i) {
142 cloud_src(0, i) = source_it->x;
143 cloud_src(1, i) = source_it->y;
144 cloud_src(2, i) = source_it->z;
145 ++source_it;
146
147 cloud_tgt(0, i) = target_it->x;
148 cloud_tgt(1, i) = target_it->y;
149 cloud_tgt(2, i) = target_it->z;
150 ++target_it;
151 }
152
153 // Call Umeyama directly from Eigen (PCL patched version until Eigen is released)
154 transformation_matrix = pcl::umeyama(cloud_src, cloud_tgt, false);
155 }
156 else {
157 source_it.reset();
158 target_it.reset();
159 // <cloud_src,cloud_src> is the source dataset
160 transformation_matrix.setIdentity();
161
162 Eigen::Matrix<Scalar, 4, 1> centroid_src, centroid_tgt;
163 // Estimate the centroids of source, target
164 compute3DCentroid(source_it, centroid_src);
165 compute3DCentroid(target_it, centroid_tgt);
166 source_it.reset();
167 target_it.reset();
168
169 // Subtract the centroids from source, target
170 Eigen::Matrix<Scalar, Eigen::Dynamic, Eigen::Dynamic> cloud_src_demean,
171 cloud_tgt_demean;
172 demeanPointCloud(source_it, centroid_src, cloud_src_demean);
173 demeanPointCloud(target_it, centroid_tgt, cloud_tgt_demean);
174
175 getTransformationFromCorrelation(cloud_src_demean,
176 centroid_src,
177 cloud_tgt_demean,
178 centroid_tgt,
179 transformation_matrix);
180 }
181}
182
183template <typename PointSource, typename PointTarget, typename Scalar>
184void
187 const Eigen::Matrix<Scalar, Eigen::Dynamic, Eigen::Dynamic>& cloud_src_demean,
188 const Eigen::Matrix<Scalar, 4, 1>& centroid_src,
189 const Eigen::Matrix<Scalar, Eigen::Dynamic, Eigen::Dynamic>& cloud_tgt_demean,
190 const Eigen::Matrix<Scalar, 4, 1>& centroid_tgt,
191 Matrix4& transformation_matrix) const
192{
193 transformation_matrix.setIdentity();
194
195 // Assemble the correlation matrix H = source * target'
196 Eigen::Matrix<Scalar, 3, 3> H =
197 (cloud_src_demean * cloud_tgt_demean.transpose()).topLeftCorner(3, 3);
198
199 // Compute the Singular Value Decomposition
200 Eigen::JacobiSVD<Eigen::Matrix<Scalar, 3, 3>> svd(
201 H, Eigen::ComputeFullU | Eigen::ComputeFullV);
202 Eigen::Matrix<Scalar, 3, 3> u = svd.matrixU();
203 Eigen::Matrix<Scalar, 3, 3> v = svd.matrixV();
204
205 // Compute R = V * U'
206 if (u.determinant() * v.determinant() < 0) {
207 for (int x = 0; x < 3; ++x)
208 v(x, 2) *= -1;
209 }
210
211 Eigen::Matrix<Scalar, 3, 3> R = v * u.transpose();
212
213 // Return the correct transformation
214 transformation_matrix.topLeftCorner(3, 3) = R;
215 const Eigen::Matrix<Scalar, 3, 1> Rc(R * centroid_src.head(3));
216 transformation_matrix.block(0, 3, 3, 1) = centroid_tgt.head(3) - Rc;
217}
218
219} // namespace registration
220} // namespace pcl
221
222//#define PCL_INSTANTIATE_TransformationEstimationSVD(T,U) template class PCL_EXPORTS
223// pcl::registration::TransformationEstimationSVD<T,U>;
224
225#endif /* PCL_REGISTRATION_TRANSFORMATION_ESTIMATION_SVD_HPP_ */
Iterator class for point clouds with or without given indices.
std::size_t size() const
Size of the range the iterator is going through.
PointCloud represents the base class in PCL for storing collections of 3D points.
std::size_t size() const
typename TransformationEstimation< PointSource, PointTarget, Scalar >::Matrix4 Matrix4
virtual void getTransformationFromCorrelation(const Eigen::Matrix< Scalar, Eigen::Dynamic, Eigen::Dynamic > &cloud_src_demean, const Eigen::Matrix< Scalar, 4, 1 > &centroid_src, const Eigen::Matrix< Scalar, Eigen::Dynamic, Eigen::Dynamic > &cloud_tgt_demean, const Eigen::Matrix< Scalar, 4, 1 > &centroid_tgt, Matrix4 &transformation_matrix) const
Obtain a 4x4 rigid transformation matrix from a correlation matrix H = src.
void estimateRigidTransformation(const pcl::PointCloud< PointSource > &cloud_src, const pcl::PointCloud< PointTarget > &cloud_tgt, Matrix4 &transformation_matrix) const override
Estimate a rigid rotation transformation between a source and a target point cloud using SVD.
void demeanPointCloud(ConstCloudIterator< PointT > &cloud_iterator, const Eigen::Matrix< Scalar, 4, 1 > &centroid, pcl::PointCloud< PointT > &cloud_out, int npts=0)
Subtract a centroid from a point cloud and return the de-meaned representation.
Definition centroid.hpp:627
unsigned int compute3DCentroid(ConstCloudIterator< PointT > &cloud_iterator, Eigen::Matrix< Scalar, 4, 1 > &centroid)
Compute the 3D (X-Y-Z) centroid of a set of points and return it as a 3D vector.
Definition centroid.hpp:56
std::vector< pcl::Correspondence, Eigen::aligned_allocator< pcl::Correspondence > > Correspondences
IndicesAllocator<> Indices
Type used for indices in PCL.
Definition types.h:133
Eigen::internal::umeyama_transform_matrix_type< Derived, OtherDerived >::type umeyama(const Eigen::MatrixBase< Derived > &src, const Eigen::MatrixBase< OtherDerived > &dst, bool with_scaling=false)
Returns the transformation between two point sets.
Definition eigen.hpp:660