otbReciprocalHAlphaImageFilter.h

Go to the documentation of this file.

/*
 * Copyright (C) 2005-2022 Centre National d'Etudes Spatiales (CNES)
 *
 * This file is part of Orfeo Toolbox
 *
 *     https://www.orfeo-toolbox.org/
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
 
 
#ifndef otbReciprocalHAlphaImageFilter_h
#define otbReciprocalHAlphaImageFilter_h
 
#include "otbMath.h"
#include "vnl/algo/vnl_complex_eigensystem.h"
#include <algorithm>
#include <vector>
#include <complex>
 
#include "otbFunctorImageFilter.h"
 
namespace otb
{
 
namespace Functor
{
 
template <class TInput, class TOutput>
class ReciprocalHAlphaFunctor
{
public:
  typedef typename std::complex<double> ComplexType;
  typedef vnl_matrix<ComplexType>       VNLMatrixType;
  typedef vnl_vector<ComplexType>       VNLVectorType;
  typedef vnl_vector<double>            VNLDoubleVectorType;
  typedef std::vector<double>           VectorType;
  typedef typename TOutput::ValueType   OutputValueType;
 
 
  inline void operator()(TOutput& result, const TInput& Coherency) const
  {
    const double T0 = static_cast<double>(Coherency[0].real());
    const double T1 = static_cast<double>(Coherency[3].real());
    const double T2 = static_cast<double>(Coherency[5].real());
 
    VNLMatrixType vnlMat(3, 3, 0.);
    vnlMat[0][0] = ComplexType(T0, 0.);
    vnlMat[0][1] = ComplexType(Coherency[1]);
    vnlMat[0][2] = ComplexType(Coherency[2]);
    vnlMat[1][0] = std::conj(ComplexType(Coherency[1]));
    vnlMat[1][1] = ComplexType(T1, 0.);
    vnlMat[1][2] = ComplexType(Coherency[4]);
    vnlMat[2][0] = std::conj(ComplexType(Coherency[2]));
    vnlMat[2][1] = std::conj(ComplexType(Coherency[4]));
    vnlMat[2][2] = ComplexType(T2, 0.);
 
    // Only compute the left symmetry to respect the previous Hermitian Analisys code
    vnl_complex_eigensystem syst(vnlMat, false, true);
    const VNLMatrixType     eigenVectors(syst.L);
    const VNLVectorType     eigenValues(syst.W);
 
    // Entropy estimation
    double totalEigenValues(0.0);
    double p[3];
    double plog[3];
    double entropy;
    double alpha;
    double anisotropy;
 
    // Sort eigen values in decreasing order
    VectorType sortedRealEigenValues(3, eigenValues[0].real());
    sortedRealEigenValues[1] = eigenValues[1].real();
    sortedRealEigenValues[2] = eigenValues[2].real();
    std::sort(sortedRealEigenValues.begin(), sortedRealEigenValues.end());
    std::reverse(sortedRealEigenValues.begin(), sortedRealEigenValues.end());
 
    // Extract the first component of each the eigen vector sorted by eigen value decrease order
    VNLVectorType sortedGreaterEigenVector(3, eigenVectors[0][0]);
    for (unsigned int i = 0; i < 3; ++i)
    {
      if (std::abs(eigenValues[1].real() - sortedRealEigenValues[i]) < m_Epsilon)
      {
        sortedGreaterEigenVector[i] = eigenVectors[1][0];
      }
      else if (std::abs(eigenValues[2].real() - sortedRealEigenValues[i]) < m_Epsilon)
      {
        sortedGreaterEigenVector[i] = eigenVectors[2][0];
      }
    }
 
    totalEigenValues = 0.0;
    for (unsigned int k = 0; k < 3; ++k)
    {
      sortedRealEigenValues[k] = std::max(sortedRealEigenValues[k], 0.);
      totalEigenValues += sortedRealEigenValues[k];
    }
 
 
    for (unsigned int k = 0; k < 3; ++k)
    {
      p[k] = sortedRealEigenValues[k] / totalEigenValues;
 
      if (p[k] < m_Epsilon) // n=log(n)-->0 when n-->0
        plog[k] = 0.0;
      else
        plog[k] = -p[k] * log(p[k]) / log(3.0);
    }
 
    entropy = 0.0;
    for (unsigned int k = 0; k < 3; ++k)
      entropy += plog[k];
 
    // alpha estimation
    double val0, val1, val2;
    double a0, a1, a2;
 
    val0 = std::abs(sortedGreaterEigenVector[0]);
    a0   = acos(std::abs(val0)) * CONST_180_PI;
 
    val1 = std::abs(sortedGreaterEigenVector[1]);
    a1   = acos(std::abs(val1)) * CONST_180_PI;
 
    val2 = std::abs(sortedGreaterEigenVector[2]);
    a2   = acos(std::abs(val2)) * CONST_180_PI;
 
    alpha = p[0] * a0 + p[1] * a1 + p[2] * a2;
 
    // Anisotropy estimation
    anisotropy = (sortedRealEigenValues[1] - sortedRealEigenValues[2]) / (sortedRealEigenValues[1] + sortedRealEigenValues[2] + m_Epsilon);
 
    result[0] = static_cast<OutputValueType>(entropy);
    result[1] = static_cast<OutputValueType>(alpha);
    result[2] = static_cast<OutputValueType>(anisotropy);
  }
 
  constexpr size_t OutputSize(...) const
  {
    // Size of the result (entropy, alpha, anisotropy)
    return 3;
  }
 
private:
  static constexpr double m_Epsilon = 1e-6;
};
} // namespace Functor
 
template <typename TInputImage, typename TOutputImage>
using ReciprocalHAlphaImageFilter = FunctorImageFilter<Functor::ReciprocalHAlphaFunctor<typename TInputImage::PixelType, typename TOutputImage::PixelType>>;
} // end namespace otb
 
#endif