-
Notifications
You must be signed in to change notification settings - Fork 432
/
CovarianceFeaturesFilter.cpp
155 lines (132 loc) · 5.61 KB
/
CovarianceFeaturesFilter.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
/******************************************************************************
* Copyright (c) 2019, Helix Re Inc. nicolas@helix.re
*
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following
* conditions are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided
* with the distribution.
* * Neither the name of Helix Re Inc. nor the
* names of its contributors may be used to endorse or promote
* products derived from this software without specific prior
* written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
* OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
* OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY
* OF SUCH DAMAGE.
****************************************************************************/
// This is an implementation of the local feature descriptors introduced in
// WEAKLY SUPERVISED SEGMENTATION-AIDED CLASSIFICATION OF URBANSCENES FROM 3D LIDAR POINT CLOUDS
// Stéphane Guinard, Loïc Landrieu, 2017
#include "CovarianceFeaturesFilter.hpp"
#include <pdal/EigenUtils.hpp>
#include <pdal/KDIndex.hpp>
#include <pdal/util/ProgramArgs.hpp>
#include <Eigen/Dense>
#include <string>
#include <vector>
#include <cmath>
namespace pdal
{
static StaticPluginInfo const s_info
{
"filters.covariancefeatures",
"Filter that calculates local features based on the covariance matrix of a point's neighborhood.",
"http://pdal.io/stages/filters.covariancefeatures.html"
};
CREATE_STATIC_STAGE(CovarianceFeaturesFilter, s_info)
std::string CovarianceFeaturesFilter::getName() const
{
return s_info.name;
}
void CovarianceFeaturesFilter::addArgs(ProgramArgs& args)
{
args.add("knn", "k-Nearest neighbors", m_knn, 10);
args.add("threads", "Number of threads used to run this filter", m_threads, 1);
args.add("feature_set", "Set of features to be computed", m_featureSet, "Dimensionality");
}
void CovarianceFeaturesFilter::addDimensions(PointLayoutPtr layout)
{
if (m_featureSet == "Dimensionality")
{
for (auto dim: {"Linearity", "Planarity", "Scattering", "Verticality"})
m_extraDims[dim] = layout->registerOrAssignDim(dim, Dimension::Type::Float);
}
}
void CovarianceFeaturesFilter::filter(PointView& view)
{
KD3Index& kdi = view.build3dIndex();
point_count_t nloops = view.size();
std::vector<std::thread> threadPool(m_threads);
for(int t = 0;t<m_threads;t++)
{
threadPool[t] = std::thread(std::bind(
[&](const PointId start, const PointId end, const PointId t)
{
for(PointId i = start;i<end;i++)
setDimensionality(view, i, kdi);
},
t*nloops/m_threads,(t+1)==m_threads?nloops:(t+1)*nloops/m_threads,t));
}
for (auto &t: threadPool)
t.join();
}
void CovarianceFeaturesFilter::setDimensionality(PointView &view, const PointId &id, const KD3Index &kid)
{
using namespace Eigen;
// find the k-nearest neighbors
auto ids = kid.neighbors(id, m_knn + 1);
// compute covariance of the neighborhood
auto B = eigen::computeCovariance(view, ids);
// perform the eigen decomposition
SelfAdjointEigenSolver<Matrix3f> solver(B);
if (solver.info() != Success)
throwError("Cannot perform eigen decomposition.");
// Extract eigenvalues and eigenvectors in decreasing order (largest eigenvalue first)
auto ev = solver.eigenvalues();
std::vector<float> lambda = {(std::max(ev[2],0.f)),
(std::max(ev[1],0.f)),
(std::max(ev[0],0.f))};
if (lambda[0] == 0)
throwError("Eigenvalues are all 0. Can't compute local features.");
auto eigenVectors = solver.eigenvectors();
std::vector<float> v1(3), v2(3), v3(3);
for (int i=0; i < 3; i++)
{
v1[i] = eigenVectors.col(2)(i);
v2[i] = eigenVectors.col(1)(i);
v3[i] = eigenVectors.col(0)(i);
}
float linearity = (sqrtf(lambda[0]) - sqrtf(lambda[1])) / sqrtf(lambda[0]);
float planarity = (sqrtf(lambda[1]) - sqrtf(lambda[2])) / sqrtf(lambda[0]);
float scattering = sqrtf(lambda[2]) / sqrtf(lambda[0]);
view.setField(m_extraDims["Linearity"], id, linearity);
view.setField(m_extraDims["Planarity"], id, planarity);
view.setField(m_extraDims["Scattering"], id, scattering);
std::vector<float> unary_vector(3);
float norm = 0;
for (int i=0; i <3 ; i++)
{
unary_vector[i] = lambda[0] * fabsf(v1[i]) + lambda[1] * fabsf(v2[i]) + lambda[2] * fabsf(v3[i]);
norm += unary_vector[i] * unary_vector[i];
}
norm = sqrtf(norm);
view.setField(m_extraDims["Verticality"], id, unary_vector[2] / norm);
}
}