Visual Servoing Platform  version 3.2.0 under development (2019-01-22)
testTukeyEstimator.cpp

Test Tukey M-Estimator.

/****************************************************************************
*
* ViSP, open source Visual Servoing Platform software.
* Copyright (C) 2005 - 2019 by Inria. All rights reserved.
*
* This software is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
* See the file LICENSE.txt at the root directory of this source
* distribution for additional information about the GNU GPL.
*
* For using ViSP with software that can not be combined with the GNU
* GPL, please contact Inria about acquiring a ViSP Professional
* Edition License.
*
* See http://visp.inria.fr for more information.
*
* This software was developed at:
* Inria Rennes - Bretagne Atlantique
* Campus Universitaire de Beaulieu
* 35042 Rennes Cedex
* France
*
* If you have questions regarding the use of this file, please contact
* Inria at visp@inria.fr
*
* This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
* WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
*
* Description:
* Test Tukey M-Estimator.
*
*****************************************************************************/
#include <cstdlib>
#include <iostream>
#include <time.h>
#include <visp3/core/vpConfig.h>
#include <visp3/core/vpGaussRand.h>
#include <visp3/core/vpRobust.h>
#include <visp3/mbt/vpMbtTukeyEstimator.h>
int main(int /*argc*/, const char ** /*argv*/)
{
size_t nb_elements = 1000;
int nb_iterations = 100;
double stdev = 0.5, mean = 0.0, noise_threshold = 1e-3;
vpGaussRand noise(stdev, mean);
noise.seed((unsigned int)time(NULL));
vpColVector residues_col((unsigned int)nb_elements);
vpColVector weights_col((unsigned int)nb_elements, 1.0), weights_col_save;
for (size_t i = 0; i < nb_elements; i++) {
residues_col[(unsigned int)i] = noise();
}
vpRobust robust((unsigned int)nb_elements);
robust.setThreshold(noise_threshold);
double t_robust = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
robust.MEstimator(vpRobust::TUKEY, residues_col, weights_col);
}
t_robust = vpTime::measureTimeMs() - t_robust;
{
vpMbtTukeyEstimator<double> tukey_estimator;
std::vector<double> residues(nb_elements);
for (size_t i = 0; i < residues.size(); i++) {
residues[i] = residues_col[(unsigned int)i];
}
std::vector<double> weights(nb_elements, 1);
double t = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
tukey_estimator.MEstimator(residues, weights, noise_threshold);
}
std::cout << "t_robust=" << t_robust << " ms ; t (double)=" << t << " ; ratio=" << (t_robust / t) << std::endl;
for (size_t i = 0; i < weights.size(); i++) {
if (!vpMath::equal(weights[i], weights_col[(unsigned int)i], noise_threshold)) {
std::cerr << "Difference between vpRobust::TUKEY and "
"vpMbtTukeyEstimator (double)!"
<< std::endl;
std::cerr << "weights_col[" << i << "]=" << weights_col[(unsigned int)i] << std::endl;
std::cerr << "weights[" << i << "]=" << weights[i] << std::endl;
return EXIT_FAILURE;
}
}
}
// Generate again for weights != 1
for (size_t i = 0; i < nb_elements; i++) {
residues_col[(unsigned int)i] = noise();
}
weights_col_save = weights_col;
t_robust = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
robust.MEstimator(vpRobust::TUKEY, residues_col, weights_col);
}
t_robust = vpTime::measureTimeMs() - t_robust;
{
vpMbtTukeyEstimator<float> tukey_estimator;
std::vector<float> residues(nb_elements);
std::vector<float> weights(nb_elements);
for (size_t i = 0; i < residues.size(); i++) {
residues[i] = (float)residues_col[(unsigned int)i];
weights[i] = (float)weights_col_save[(unsigned int)i];
}
double t = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
tukey_estimator.MEstimator(residues, weights, (float)noise_threshold);
}
std::cout << "t_robust=" << t_robust << " ms ; t (float)=" << t << " ; ratio=" << (t_robust / t) << std::endl;
for (size_t i = 0; i < weights.size(); i++) {
if (!vpMath::equal(weights[i], weights_col[(unsigned int)i], noise_threshold)) {
std::cerr << "Difference between vpRobust::TUKEY and "
"vpMbtTukeyEstimator (float)!"
<< std::endl;
std::cerr << "weights_col[" << i << "]=" << weights_col[(unsigned int)i] << std::endl;
std::cerr << "weights[" << i << "]=" << weights[i] << std::endl;
return EXIT_FAILURE;
}
}
}
// Generate again for weights != 1 and vpColVector type
for (size_t i = 0; i < nb_elements; i++) {
residues_col[(unsigned int)i] = noise();
}
weights_col_save = weights_col;
t_robust = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
robust.MEstimator(vpRobust::TUKEY, residues_col, weights_col);
}
t_robust = vpTime::measureTimeMs() - t_robust;
{
vpMbtTukeyEstimator<double> tukey_estimator;
vpColVector residues = residues_col;
vpColVector weights = weights_col_save;
double t = vpTime::measureTimeMs();
for (int i = 0; i < nb_iterations; i++) {
tukey_estimator.MEstimator(residues, weights, noise_threshold);
}
std::cout << "t_robust=" << t_robust << " ms ; t (vpColVector)=" << t << " ; ratio=" << (t_robust / t) << std::endl;
for (size_t i = 0; i < weights.size(); i++) {
if (!vpMath::equal(weights[(unsigned int)i], weights_col[(unsigned int)i], noise_threshold)) {
std::cerr << "Difference between vpRobust::TUKEY and "
"vpMbtTukeyEstimator (float)!"
<< std::endl;
std::cerr << "weights_col[" << i << "]=" << weights_col[(unsigned int)i] << std::endl;
std::cerr << "weights[" << i << "]=" << weights[(unsigned int)i] << std::endl;
return EXIT_FAILURE;
}
}
}
std::cout << "vpMbtTukeyEstimator returns the same values than vpRobust::TUKEY." << std::endl;
return EXIT_SUCCESS;
}