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
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
|
// Copyright (c) 2021, Viktor Larsson
// 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 the copyright holder 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 COPYRIGHT HOLDERS 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.
#include "absolute_pose.h"
#include "PoseLib/robust/bundle.h"
#include "PoseLib/solvers/gp3p.h"
#include "PoseLib/solvers/p1p2ll.h"
#include "PoseLib/solvers/p2p1ll.h"
#include "PoseLib/solvers/p3ll.h"
#include "PoseLib/solvers/p3p.h"
#include "PoseLib/solvers/p5lp_radial.h"
namespace poselib {
void AbsolutePoseEstimator::generate_models(std::vector<CameraPose> *models) {
sampler.generate_sample(&sample);
for (size_t k = 0; k < sample_sz; ++k) {
xs[k] = x[sample[k]].homogeneous().normalized();
Xs[k] = X[sample[k]];
}
p3p(xs, Xs, models);
}
double AbsolutePoseEstimator::score_model(const CameraPose &pose, size_t *inlier_count) const {
return compute_msac_score(pose, x, X, opt.max_reproj_error * opt.max_reproj_error, inlier_count);
}
void AbsolutePoseEstimator::refine_model(CameraPose *pose) const {
BundleOptions bundle_opt;
bundle_opt.loss_type = BundleOptions::LossType::TRUNCATED;
bundle_opt.loss_scale = opt.max_reproj_error;
bundle_opt.max_iterations = 25;
// TODO: for high outlier scenarios, make a copy of (x,X) and find points close to inlier threshold
// TODO: experiment with good thresholds for copy vs iterating full point set
bundle_adjust(x, X, pose, bundle_opt);
}
void GeneralizedAbsolutePoseEstimator::generate_models(std::vector<CameraPose> *models) {
draw_sample(sample_sz, num_pts_camera, &sample, rng);
for (size_t k = 0; k < sample_sz; ++k) {
const size_t cam_k = sample[k].first;
const size_t pt_k = sample[k].second;
ps[k] = camera_centers[cam_k];
xs[k] = rig_poses[cam_k].derotate(x[cam_k][pt_k].homogeneous().normalized());
Xs[k] = X[cam_k][pt_k];
}
gp3p(ps, xs, Xs, models);
}
double GeneralizedAbsolutePoseEstimator::score_model(const CameraPose &pose, size_t *inlier_count) const {
const double sq_threshold = opt.max_reproj_error * opt.max_reproj_error;
double score = 0;
*inlier_count = 0;
size_t cam_inlier_count;
for (size_t k = 0; k < num_cams; ++k) {
CameraPose full_pose;
full_pose.q = quat_multiply(rig_poses[k].q, pose.q);
full_pose.t = rig_poses[k].rotate(pose.t) + rig_poses[k].t;
score += compute_msac_score(full_pose, x[k], X[k], sq_threshold, &cam_inlier_count);
*inlier_count += cam_inlier_count;
}
return score;
}
void GeneralizedAbsolutePoseEstimator::refine_model(CameraPose *pose) const {
BundleOptions bundle_opt;
bundle_opt.loss_type = BundleOptions::LossType::TRUNCATED;
bundle_opt.loss_scale = opt.max_reproj_error;
bundle_opt.max_iterations = 25;
generalized_bundle_adjust(x, X, rig_poses, pose, bundle_opt);
}
void AbsolutePosePointLineEstimator::generate_models(std::vector<CameraPose> *models) {
draw_sample(sample_sz, num_data, &sample, rng);
size_t pt_idx = 0;
size_t line_idx = 0;
for (size_t k = 0; k < sample_sz; ++k) {
size_t idx = sample[k];
if (idx < points2D.size()) {
// we sampled a point correspondence
xs[pt_idx] = points2D[idx].homogeneous();
xs[pt_idx].normalize();
Xs[pt_idx] = points3D[idx];
pt_idx++;
} else {
// we sampled a line correspondence
idx -= points2D.size();
ls[line_idx] = lines2D[idx].x1.homogeneous().cross(lines2D[idx].x2.homogeneous());
ls[line_idx].normalize();
Cs[line_idx] = lines3D[idx].X1;
Vs[line_idx] = lines3D[idx].X2 - lines3D[idx].X1;
Vs[line_idx].normalize();
line_idx++;
}
}
if (pt_idx == 3 && line_idx == 0) {
p3p(xs, Xs, models);
} else if (pt_idx == 2 && line_idx == 1) {
p2p1ll(xs, Xs, ls, Cs, Vs, models);
} else if (pt_idx == 1 && line_idx == 2) {
p1p2ll(xs, Xs, ls, Cs, Vs, models);
} else if (pt_idx == 0 && line_idx == 3) {
p3ll(ls, Cs, Vs, models);
}
}
double AbsolutePosePointLineEstimator::score_model(const CameraPose &pose, size_t *inlier_count) const {
size_t point_inliers, line_inliers;
double score_pt =
compute_msac_score(pose, points2D, points3D, opt.max_reproj_error * opt.max_reproj_error, &point_inliers);
double score_l =
compute_msac_score(pose, lines2D, lines3D, opt.max_epipolar_error * opt.max_epipolar_error, &line_inliers);
*inlier_count = point_inliers + line_inliers;
return score_pt + score_l;
}
void AbsolutePosePointLineEstimator::refine_model(CameraPose *pose) const {
BundleOptions bundle_opt;
bundle_opt.loss_type = BundleOptions::LossType::TRUNCATED;
bundle_opt.loss_scale = opt.max_reproj_error;
bundle_opt.max_iterations = 25;
BundleOptions line_bundle_opt;
line_bundle_opt.loss_type = BundleOptions::LossType::TRUNCATED;
line_bundle_opt.loss_scale = opt.max_epipolar_error;
bundle_adjust(points2D, points3D, lines2D, lines3D, pose, bundle_opt, line_bundle_opt);
}
void Radial1DAbsolutePoseEstimator::generate_models(std::vector<CameraPose> *models) {
sampler.generate_sample(&sample);
for (size_t k = 0; k < sample_sz; ++k) {
xs[k] = x[sample[k]].normalized();
Xs[k] = X[sample[k]];
}
p5lp_radial(xs, Xs, models);
}
double Radial1DAbsolutePoseEstimator::score_model(const CameraPose &pose, size_t *inlier_count) const {
return compute_msac_score_1D_radial(pose, x, X, opt.max_reproj_error * opt.max_reproj_error, inlier_count);
}
void Radial1DAbsolutePoseEstimator::refine_model(CameraPose *pose) const {
BundleOptions bundle_opt;
bundle_opt.loss_type = BundleOptions::LossType::TRUNCATED;
bundle_opt.loss_scale = opt.max_reproj_error;
bundle_opt.max_iterations = 25;
// TODO: for high outlier scenarios, make a copy of (x,X) and find points close to inlier threshold
// TODO: experiment with good thresholds for copy vs iterating full point set
bundle_adjust_1D_radial(x, X, pose, bundle_opt);
}
} // namespace poselib
|
