distances.h

/*
 * Software License Agreement (BSD License)
 *
 *  Point Cloud Library (PCL) - www.pointclouds.org
 *  Copyright (c) 2010-2011, Willow Garage, Inc.
 *  Copyright (c) 2012-, Open Perception, Inc.
 *
 *  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(s) 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.
 *
 * $Id$
 *
 */
 
#pragma once
 
#include <Eigen/Core>
 
#include <algorithm>
#include <cstring>
#include <vector>
 
namespace pcl {
namespace distances {
 
/** \brief Compute the median value from a set of doubles
 * \param[in] fvec the set of doubles
 * \param[in] m the number of doubles in the set
 */
inline double
computeMedian(double* fvec, int m)
{
  // Copy the values to vectors for faster sorting
  std::vector<double> data(fvec, fvec + m);
 
  std::nth_element(data.begin(), data.begin() + (data.size() >> 1), data.end());
  return (data[data.size() >> 1]);
}
 
/** \brief Use a Huber kernel to estimate the distance between two vectors
 * \param[in] p_src the first eigen vector
 * \param[in] p_tgt the second eigen vector
 * \param[in] sigma the sigma value
 */
inline double
huber(const Eigen::Vector4f& p_src, const Eigen::Vector4f& p_tgt, double sigma)
{
  Eigen::Array4f diff = (p_tgt.array() - p_src.array()).abs();
  double norm = 0.0;
  for (int i = 0; i < 3; ++i) {
    if (diff[i] < sigma)
      norm += diff[i] * diff[i];
    else
      norm += 2.0 * sigma * diff[i] - sigma * sigma;
  }
  return (norm);
}
 
/** \brief Use a Huber kernel to estimate the distance between two vectors
 * \param[in] diff the norm difference between two vectors
 * \param[in] sigma the sigma value
 */
inline double
huber(double diff, double sigma)
{
  double norm = 0.0;
  if (diff < sigma)
    norm += diff * diff;
  else
    norm += 2.0 * sigma * diff - sigma * sigma;
  return (norm);
}
 
/** \brief Use a Gedikli kernel to estimate the distance between two vectors
 * (for more information, see
 * \param[in] val the norm difference between two vectors
 * \param[in] clipping the clipping value
 * \param[in] slope the slope. Default: 4
 */
inline double
gedikli(double val, double clipping, double slope = 4)
{
  return (1.0 / (1.0 + pow(std::abs(val) / clipping, slope)));
}
 
/** \brief Compute the Manhattan distance between two eigen vectors.
 * \param[in] p_src the first eigen vector
 * \param[in] p_tgt the second eigen vector
 */
inline double
l1(const Eigen::Vector4f& p_src, const Eigen::Vector4f& p_tgt)
{
  return ((p_src.array() - p_tgt.array()).abs().sum());
}
 
/** \brief Compute the Euclidean distance between two eigen vectors.
 * \param[in] p_src the first eigen vector
 * \param[in] p_tgt the second eigen vector
 */
inline double
l2(const Eigen::Vector4f& p_src, const Eigen::Vector4f& p_tgt)
{
  return ((p_src - p_tgt).norm());
}
 
/** \brief Compute the squared Euclidean distance between two eigen vectors.
 * \param[in] p_src the first eigen vector
 * \param[in] p_tgt the second eigen vector
 */
inline double
l2Sqr(const Eigen::Vector4f& p_src, const Eigen::Vector4f& p_tgt)
{
  return ((p_src - p_tgt).squaredNorm());
}
} // namespace distances
} // namespace pcl