itkSampleClassifierWithMask.txx

Go to the documentation of this file.

/*=========================================================================

  Program:   Insight Segmentation & Registration Toolkit
  Module:    $RCSfile: itkSampleClassifierWithMask.txx,v $
  Language:  C++
  Date:      $Date: 2009-03-04 19:29:54 $
  Version:   $Revision: 1.9 $

  Copyright (c) Insight Software Consortium. All rights reserved.
  See ITKCopyright.txt or http://www.itk.org/HTML/Copyright.htm for details.

     This software is distributed WITHOUT ANY WARRANTY; without even 
     the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR 
     PURPOSE.  See the above copyright notices for more information.

=========================================================================*/
#ifndef __itkSampleClassifierWithMask_txx
#define __itkSampleClassifierWithMask_txx

#include "itkSampleClassifierWithMask.h"

namespace itk { 
namespace Statistics {

template< class TSample, class TMaskSample >
SampleClassifierWithMask< TSample, TMaskSample >
::SampleClassifierWithMask()
{
  m_OtherClassLabel = 0;
  m_Mask = 0;
}

template< class TSample, class TMaskSample >
void
SampleClassifierWithMask< TSample, TMaskSample >
::PrintSelf(std::ostream& os, Indent indent) const
{
  Superclass::PrintSelf(os,indent);

  os << indent << "Mask: ";
  if ( m_Mask.IsNotNull() )
    {
    os << m_Mask << std::endl;
    }
  else
    {
    os << "not set." << std::endl;
    }

  os << indent << "SelectedClassLabels: ";
  for ( unsigned int i = 0; i < m_SelectedClassLabels.size(); ++i )
    {
    os << " " << m_SelectedClassLabels[i];
    }
  os << std::endl;
  os << indent << "OtherClassLabel: " << m_OtherClassLabel << std::endl;
}

template< class TSample, class TMaskSample >
void
SampleClassifierWithMask< TSample, TMaskSample >
::SetMask(TMaskSample* mask)
{
  if ( m_Mask != mask )
    {
    m_Mask = mask;
    }
}

template< class TSample, class TMaskSample >
void
SampleClassifierWithMask< TSample, TMaskSample >
::GenerateData()
{
  unsigned int i;
  typename TSample::ConstIterator iter = this->GetSample()->Begin();
  typename TSample::ConstIterator end = this->GetSample()->End();
  typename TSample::MeasurementVectorType measurements;

  typename TMaskSample::Iterator m_iter = this->GetMask()->Begin();

  OutputType* output = this->GetOutput();
  output->Resize(this->GetSample()->Size());
  std::vector< double > discriminantScores;
  unsigned int numberOfClasses = this->GetNumberOfClasses();
  discriminantScores.resize(numberOfClasses);
  output->SetNumberOfClasses(numberOfClasses + 1);
  unsigned int classLabel;
  typename Superclass::DecisionRuleType::Pointer rule = 
    this->GetDecisionRule();
  typename Superclass::ClassLabelVectorType classLabels = 
    this->GetMembershipFunctionClassLabels();
  
  if ( this->GetMask()->Size() != this->GetSample()->Size() )
    {
    itkExceptionMacro("The sizes of the mask sample and the input sample do not match.");
    }

  if ( classLabels.size() != this->GetNumberOfMembershipFunctions() )
    {
    while (iter != end)
      {
      measurements = iter.GetMeasurementVector();
      if ( std::find(m_SelectedClassLabels.begin(), 
                     m_SelectedClassLabels.end(), 
                     m_iter.GetMeasurementVector()[0]) != 
           m_SelectedClassLabels.end() )
        {
        for (i = 0; i < numberOfClasses; i++)
          {
          discriminantScores[i] = 
            (this->GetMembershipFunction(i))->Evaluate(measurements);
          }
        classLabel = rule->Evaluate(discriminantScores);
        }
      else
        {
        classLabel = m_OtherClassLabel;
        }
      output->AddInstance(classLabel, iter.GetInstanceIdentifier());
      ++iter;
      ++m_iter;
      }
    }
  else
    {
    while (iter != end)
      {
      measurements = iter.GetMeasurementVector();
      if ( std::find(m_SelectedClassLabels.begin(), 
                     m_SelectedClassLabels.end(), 
                     m_iter.GetMeasurementVector()[0]) != 
           m_SelectedClassLabels.end() )
        {
        for (i = 0; i < numberOfClasses; i++)
          {
          discriminantScores[i] =
            (this->GetMembershipFunction(i))->Evaluate(measurements);
          }
        classLabel = rule->Evaluate(discriminantScores);
        output->AddInstance(classLabels[classLabel], 
                            iter.GetInstanceIdentifier());
        }
      else
        {
        output->AddInstance(m_OtherClassLabel, 
                            iter.GetInstanceIdentifier());
        }
      ++iter;
      ++m_iter;
      }
    }
}

} // end of namespace Statistics 
} // end of namespace itk

#endif