ceres-solver-v1 / colmap /src /estimators /absolute_pose_test.cc
camenduru's picture
ceres-solver and colmap
7b7496d
// Copyright (c) 2022, ETH Zurich and UNC Chapel Hill.
// 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 ETH Zurich and UNC Chapel Hill 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 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.
//
// Author: Johannes L. Schoenberger (jsch-at-demuc-dot-de)
#define TEST_NAME "base/absolute_pose"
#include "util/testing.h"
#include <Eigen/Core>
#include "base/pose.h"
#include "base/similarity_transform.h"
#include "estimators/absolute_pose.h"
#include "estimators/essential_matrix.h"
#include "optim/ransac.h"
#include "util/random.h"
using namespace colmap;
BOOST_AUTO_TEST_CASE(TestP3P) {
SetPRNGSeed(0);
std::vector<Eigen::Vector3d> points3D;
points3D.emplace_back(1, 1, 1);
points3D.emplace_back(0, 1, 1);
points3D.emplace_back(3, 1.0, 4);
points3D.emplace_back(3, 1.1, 4);
points3D.emplace_back(3, 1.2, 4);
points3D.emplace_back(3, 1.3, 4);
points3D.emplace_back(3, 1.4, 4);
points3D.emplace_back(2, 1, 7);
auto points3D_faulty = points3D;
for (size_t i = 0; i < points3D.size(); ++i) {
points3D_faulty[i](0) = 20;
}
for (double qx = 0; qx < 1; qx += 0.2) {
for (double tx = 0; tx < 1; tx += 0.1) {
const SimilarityTransform3 orig_tform(1, Eigen::Vector4d(1, qx, 0, 0),
Eigen::Vector3d(tx, 0, 0));
// Project points to camera coordinate system.
std::vector<Eigen::Vector2d> points2D;
for (size_t i = 0; i < points3D.size(); ++i) {
Eigen::Vector3d point3D_camera = points3D[i];
orig_tform.TransformPoint(&point3D_camera);
points2D.push_back(point3D_camera.hnormalized());
}
RANSACOptions options;
options.max_error = 1e-5;
RANSAC<P3PEstimator> ransac(options);
const auto report = ransac.Estimate(points2D, points3D);
BOOST_CHECK_EQUAL(report.success, true);
// Test if correct transformation has been determined.
const double matrix_diff =
(orig_tform.Matrix().topLeftCorner<3, 4>() - report.model).norm();
BOOST_CHECK(matrix_diff < 1e-2);
// Test residuals of exact points.
std::vector<double> residuals;
P3PEstimator::Residuals(points2D, points3D, report.model, &residuals);
for (size_t i = 0; i < residuals.size(); ++i) {
BOOST_CHECK(residuals[i] < 1e-3);
}
// Test residuals of faulty points.
P3PEstimator::Residuals(points2D, points3D_faulty, report.model,
&residuals);
for (size_t i = 0; i < residuals.size(); ++i) {
BOOST_CHECK(residuals[i] > 0.1);
}
}
}
}
BOOST_AUTO_TEST_CASE(TestEPNP) {
SetPRNGSeed(0);
std::vector<Eigen::Vector3d> points3D;
points3D.emplace_back(1, 1, 1);
points3D.emplace_back(0, 1, 1);
points3D.emplace_back(3, 1.0, 4);
points3D.emplace_back(3, 1.1, 4);
points3D.emplace_back(3, 1.2, 4);
points3D.emplace_back(3, 1.3, 4);
points3D.emplace_back(3, 1.4, 4);
points3D.emplace_back(2, 1, 7);
auto points3D_faulty = points3D;
for (size_t i = 0; i < points3D.size(); ++i) {
points3D_faulty[i](0) = 20;
}
for (double qx = 0; qx < 1; qx += 0.2) {
for (double tx = 0; tx < 1; tx += 0.1) {
const SimilarityTransform3 orig_tform(1, Eigen::Vector4d(1, qx, 0, 0),
Eigen::Vector3d(tx, 0, 0));
// Project points to camera coordinate system.
std::vector<Eigen::Vector2d> points2D;
for (size_t i = 0; i < points3D.size(); ++i) {
Eigen::Vector3d point3D_camera = points3D[i];
orig_tform.TransformPoint(&point3D_camera);
points2D.push_back(point3D_camera.hnormalized());
}
RANSACOptions options;
options.max_error = 1e-5;
RANSAC<EPNPEstimator> ransac(options);
const auto report = ransac.Estimate(points2D, points3D);
BOOST_CHECK_EQUAL(report.success, true);
// Test if correct transformation has been determined.
const double matrix_diff =
(orig_tform.Matrix().topLeftCorner<3, 4>() - report.model).norm();
BOOST_CHECK(matrix_diff < 1e-3);
// Test residuals of exact points.
std::vector<double> residuals;
EPNPEstimator::Residuals(points2D, points3D, report.model, &residuals);
for (size_t i = 0; i < residuals.size(); ++i) {
BOOST_CHECK(residuals[i] < 1e-3);
}
// Test residuals of faulty points.
EPNPEstimator::Residuals(points2D, points3D_faulty, report.model,
&residuals);
for (size_t i = 0; i < residuals.size(); ++i) {
BOOST_CHECK(residuals[i] > 0.1);
}
}
}
}
BOOST_AUTO_TEST_CASE(TestEPNP_BrokenSolveSignCase) {
std::vector<Eigen::Vector2d> points2D;
points2D.emplace_back(-2.6783007931074532e-01, 5.3457197430746251e-01);
points2D.emplace_back(-4.2629907287470264e-01, 7.5623350319519789e-01);
points2D.emplace_back(-1.6767413005963930e-01, -1.3387172544910089e-01);
points2D.emplace_back(-5.6616329720373559e-02, 2.3621156497739373e-01);
points2D.emplace_back(-1.7721225948969935e-01, 2.3395366792735982e-02);
points2D.emplace_back(-5.1836259886632222e-02, -4.4380694271927049e-02);
points2D.emplace_back(-3.5897765845560037e-01, 1.6252721078589397e-01);
points2D.emplace_back(2.7057324473684058e-01, -1.4067450104631887e-01);
points2D.emplace_back(-2.5811166424334520e-01, 8.0167171300227366e-02);
points2D.emplace_back(2.0239567448222310e-02, -3.2845953375344145e-01);
points2D.emplace_back(4.2571014715170657e-01, -2.8321173570154773e-01);
points2D.emplace_back(-5.4597596412987237e-01, 9.1431935871671977e-02);
std::vector<Eigen::Vector3d> points3D;
points3D.emplace_back(4.4276865308679305e+00, -1.3384364366019632e+00,
-3.5997423085253892e+00);
points3D.emplace_back(2.7278555252512309e+00, -3.8152996187231392e-01,
-2.6558518399902824e+00);
points3D.emplace_back(4.8548566083054894e+00, -1.4756197433631739e+00,
-6.8274946022490501e-01);
points3D.emplace_back(3.1523013527998449e+00, -1.3377020437938025e+00,
-1.6443269301929087e+00);
points3D.emplace_back(3.8551679771512073e+00, -1.0557700545885551e+00,
-1.1695994508851486e+00);
points3D.emplace_back(5.9571373150353812e+00, -2.6120646101684555e+00,
-1.0841441206050342e+00);
points3D.emplace_back(6.3287088499358894e+00, -1.1761274755817175e+00,
-2.5951879774151583e+00);
points3D.emplace_back(2.3005305990121250e+00, -1.4019796626800123e+00,
-4.4485464455072321e-01);
points3D.emplace_back(5.9816859934587354e+00, -1.4211814511691452e+00,
-2.0285923889293449e+00);
points3D.emplace_back(5.2543344690665457e+00, -2.3389255564264144e+00,
4.3708173185524052e-01);
points3D.emplace_back(3.2181599245991688e+00, -2.8906671988445098e+00,
2.6825718150064348e-01);
points3D.emplace_back(4.4592895306946758e+00, -9.1235241641579902e-03,
-1.6555237117970871e+00);
const std::vector<EPNPEstimator::M_t> output =
EPNPEstimator::Estimate(points2D, points3D);
BOOST_CHECK_EQUAL(output.size(), 1);
double reproj = 0.0;
for (size_t i = 0; i < points3D.size(); ++i) {
reproj +=
((output[0] * points3D[i].homogeneous()).hnormalized() - points2D[i])
.norm();
}
BOOST_CHECK(reproj < 0.2);
}