ia_ransac.h

Go to the documentation of this file.

/*
 * Software License Agreement (BSD License)
 *
 *  Point Cloud Library (PCL) - www.pointclouds.org
 *  Copyright (c) 2010-2012, 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.
 *
 * $Id$
 *
 */
#ifndef IA_RANSAC_H_
#define IA_RANSAC_H_

#include <pcl/registration/registration.h>
#include <pcl/registration/transformation_estimation_svd.h>

namespace pcl
{
  template <typename PointSource, typename PointTarget, typename FeatureT>
  class SampleConsensusInitialAlignment : public Registration<PointSource, PointTarget>
  {
    public:
      using Registration<PointSource, PointTarget>::reg_name_;
      using Registration<PointSource, PointTarget>::input_;
      using Registration<PointSource, PointTarget>::indices_;
      using Registration<PointSource, PointTarget>::target_;
      using Registration<PointSource, PointTarget>::final_transformation_;
      using Registration<PointSource, PointTarget>::transformation_;
      using Registration<PointSource, PointTarget>::corr_dist_threshold_;
      using Registration<PointSource, PointTarget>::min_number_correspondences_;
      using Registration<PointSource, PointTarget>::max_iterations_;
      using Registration<PointSource, PointTarget>::tree_;
      using Registration<PointSource, PointTarget>::transformation_estimation_;
      using Registration<PointSource, PointTarget>::getClassName;

      typedef typename Registration<PointSource, PointTarget>::PointCloudSource PointCloudSource;
      typedef typename PointCloudSource::Ptr PointCloudSourcePtr;
      typedef typename PointCloudSource::ConstPtr PointCloudSourceConstPtr;

      typedef typename Registration<PointSource, PointTarget>::PointCloudTarget PointCloudTarget;

      typedef PointIndices::Ptr PointIndicesPtr;
      typedef PointIndices::ConstPtr PointIndicesConstPtr;

      typedef pcl::PointCloud<FeatureT> FeatureCloud;
      typedef typename FeatureCloud::Ptr FeatureCloudPtr;
      typedef typename FeatureCloud::ConstPtr FeatureCloudConstPtr;

      typedef boost::shared_ptr<SampleConsensusInitialAlignment<PointSource, PointTarget, FeatureT> > Ptr;
      typedef boost::shared_ptr<const SampleConsensusInitialAlignment<PointSource, PointTarget, FeatureT> > ConstPtr;


      class ErrorFunctor
      {
        public:
          virtual ~ErrorFunctor () {}
          virtual float operator () (float d) const = 0;
      };

      class HuberPenalty : public ErrorFunctor
      {
        private:
          HuberPenalty () {}
        public:
          HuberPenalty (float threshold)  : threshold_ (threshold) {}
          virtual float operator () (float e) const
          { 
            if (e <= threshold_)
              return (0.5 * e*e); 
            else
              return (0.5 * threshold_ * (2.0 * fabs (e) - threshold_));
          }
        protected:
          float threshold_;
      };

      class TruncatedError : public ErrorFunctor
      {
        private:
          TruncatedError () {}
        public:
          virtual ~TruncatedError () {}

          TruncatedError (float threshold) : threshold_ (threshold) {}
          virtual float operator () (float e) const
          { 
            if (e <= threshold_)
              return (e / threshold_);
            else
              return (1.0);
          }
        protected:
          float threshold_;
      };

      typedef typename KdTreeFLANN<FeatureT>::Ptr FeatureKdTreePtr; 
      SampleConsensusInitialAlignment () : 
        input_features_ (), target_features_ (), 
        nr_samples_(3), min_sample_distance_ (0.0f), k_correspondences_ (10), 
        feature_tree_ (new pcl::KdTreeFLANN<FeatureT>),
        error_functor_ ()
      {
        reg_name_ = "SampleConsensusInitialAlignment";
        max_iterations_ = 1000;

        // Setting a non-std::numeric_limits<double>::max () value to corr_dist_threshold_ to make it play nicely with TruncatedError
        corr_dist_threshold_ = 100.0f;
        transformation_estimation_.reset (new pcl::registration::TransformationEstimationSVD<PointSource, PointTarget>);
      };

      void 
      setSourceFeatures (const FeatureCloudConstPtr &features);

      inline FeatureCloudConstPtr const 
      getSourceFeatures () { return (input_features_); }

      void 
      setTargetFeatures (const FeatureCloudConstPtr &features);

      inline FeatureCloudConstPtr const 
      getTargetFeatures () { return (target_features_); }

      void 
      setMinSampleDistance (float min_sample_distance) { min_sample_distance_ = min_sample_distance; }

      float 
      getMinSampleDistance () { return (min_sample_distance_); }

      void 
      setNumberOfSamples (int nr_samples) { nr_samples_ = nr_samples; }

      int 
      getNumberOfSamples () { return (nr_samples_); }

      void
      setCorrespondenceRandomness (int k) { k_correspondences_ = k; }

      int
      getCorrespondenceRandomness () { return (k_correspondences_); }

      void
      setErrorFunction (const boost::shared_ptr<ErrorFunctor> & error_functor) { error_functor_ = error_functor; }

      boost::shared_ptr<ErrorFunctor>
      getErrorFunction () { return (error_functor_); }

    protected:
      inline int 
      getRandomIndex (int n) { return (static_cast<int> (n * (rand () / (RAND_MAX + 1.0)))); };
      
      void 
      selectSamples (const PointCloudSource &cloud, int nr_samples, float min_sample_distance, 
                     std::vector<int> &sample_indices);

      void 
      findSimilarFeatures (const FeatureCloud &input_features, const std::vector<int> &sample_indices, 
                           std::vector<int> &corresponding_indices);

      float 
      computeErrorMetric (const PointCloudSource &cloud, float threshold);

      virtual void 
      computeTransformation (PointCloudSource &output, const Eigen::Matrix4f& guess);

      FeatureCloudConstPtr input_features_;

      FeatureCloudConstPtr target_features_;  

      int nr_samples_;

      float min_sample_distance_;

      int k_correspondences_;
     
      FeatureKdTreePtr feature_tree_;               

      boost::shared_ptr<ErrorFunctor> error_functor_;
    public:
      EIGEN_MAKE_ALIGNED_OPERATOR_NEW
  };
}

#include <pcl/registration/impl/ia_ransac.hpp>

#endif  //#ifndef IA_RANSAC_H_