1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
|
/*=========================================================================
Program: Advanced Normalization Tools
Copyright (c) ConsortiumOfANTS. All rights reserved.
See accompanying COPYING.txt or
https://github.com/stnava/ANTs/blob/master/ANTSCopyright.txt
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 __antsManifoldParzenWindowsListSampleFunction_hxx
#define __antsManifoldParzenWindowsListSampleFunction_hxx
namespace itk
{
namespace ants
{
namespace Statistics
{
template <typename TListSample, typename TOutput, typename TCoordRep>
ManifoldParzenWindowsListSampleFunction<TListSample, TOutput, TCoordRep>::ManifoldParzenWindowsListSampleFunction()
{
this->m_KdTreeGenerator = nullptr;
this->m_EvaluationKNeighborhood = 50;
this->m_RegularizationSigma = 1.0;
this->m_CovarianceKNeighborhood = 0;
this->m_KernelSigma = 0.0;
}
template <typename TListSample, typename TOutput, typename TCoordRep>
ManifoldParzenWindowsListSampleFunction<TListSample, TOutput, TCoordRep>::~ManifoldParzenWindowsListSampleFunction() =
default;
template <typename TListSample, typename TOutput, typename TCoordRep>
void
ManifoldParzenWindowsListSampleFunction<TListSample, TOutput, TCoordRep>::SetInputListSample(
const InputListSampleType * ptr)
{
Superclass::SetInputListSample(ptr);
if (!this->GetInputListSample())
{
return;
}
if (this->GetInputListSample()->Size() <= 1)
{
itkWarningMacro("The input list sample has <= 1 element."
<< "Function evaluations will be equal to 0.");
return;
}
/**
* Generate KdTree and create set of gaussians from input point set
*/
this->m_KdTreeGenerator = TreeGeneratorType::New();
this->m_KdTreeGenerator->SetSample(const_cast<InputListSampleType *>(this->GetInputListSample()));
this->m_KdTreeGenerator->SetBucketSize(16);
this->m_KdTreeGenerator->Update();
/**
* Calculate covariance matrices
*/
this->m_Gaussians.resize(this->GetInputListSample()->Size());
const unsigned int Dimension = this->GetInputListSample()->GetMeasurementVectorSize();
unsigned long count = 0;
typename InputListSampleType::ConstIterator It = this->GetInputListSample()->Begin();
while (It != this->GetInputListSample()->End())
{
InputMeasurementVectorType inputMeasurement = It.GetMeasurementVector();
typename GaussianType::MeanVectorType mean(Dimension);
for (unsigned int d = 0; d < Dimension; d++)
{
mean[d] = inputMeasurement[d];
}
this->m_Gaussians[count] = GaussianType::New();
this->m_Gaussians[count]->SetMean(mean);
if (this->m_CovarianceKNeighborhood > 0)
{
/** Temporarily set the covariance */
CovarianceMatrixType Cov(Dimension, Dimension);
Cov.SetIdentity();
Cov *= this->m_KernelSigma;
this->m_Gaussians[count]->SetCovariance(Cov);
Cov.Fill(0);
typename TreeGeneratorType::KdTreeType ::InstanceIdentifierVectorType neighbors;
unsigned int numberOfNeighbors =
std::min(this->m_CovarianceKNeighborhood, static_cast<unsigned int>(this->GetInputListSample()->Size()));
this->m_KdTreeGenerator->GetOutput()->Search(inputMeasurement, numberOfNeighbors, neighbors);
RealType denominator = 0.0;
for (unsigned int j = 0; j < numberOfNeighbors; j++)
{
if (neighbors[j] != count && neighbors[j] < this->GetInputListSample()->Size())
{
InputMeasurementVectorType neighbor =
this->m_KdTreeGenerator->GetOutput()->GetMeasurementVector(neighbors[j]);
RealType kernelValue = this->m_Gaussians[count]->Evaluate(neighbor);
if (this->GetListSampleWeights()->Size() == this->m_Gaussians.size())
{
kernelValue *= static_cast<RealType>((*this->GetListSampleWeights())[count]);
}
denominator += kernelValue;
if (kernelValue > NumericTraits<RealType>::ZeroValue())
{
for (unsigned int m = 0; m < Dimension; m++)
{
for (unsigned int n = m; n < Dimension; n++)
{
RealType covariance =
kernelValue * (neighbor[m] - inputMeasurement[m]) * (neighbor[n] - inputMeasurement[n]);
Cov(m, n) += static_cast<typename CovarianceMatrixType::ComponentType>(covariance);
Cov(n, m) += Cov(m, n);
}
}
}
}
}
if (denominator > NumericTraits<RealType>::ZeroValue())
{
Cov /= denominator;
}
for (unsigned int m = 0; m < Dimension; m++)
{
Cov(m, m) += static_cast<typename CovarianceMatrixType::ComponentType>(this->m_RegularizationSigma *
this->m_RegularizationSigma);
}
this->m_Gaussians[count]->SetCovariance(Cov);
}
else
{
CovarianceMatrixType Cov(Dimension, Dimension);
Cov.SetIdentity();
Cov *= this->m_RegularizationSigma;
this->m_Gaussians[count]->SetCovariance(Cov);
}
++It;
++count;
}
/**
* Calculate normalization factor
*/
this->m_NormalizationFactor = 0.0;
for (unsigned int i = 0; i < this->m_Gaussians.size(); i++)
{
if (this->GetListSampleWeights()->Size() == this->m_Gaussians.size())
{
this->m_NormalizationFactor += static_cast<RealType>((*this->GetListSampleWeights())[i]);
}
else
{
this->m_NormalizationFactor += NumericTraits<RealType>::OneValue();
}
}
}
template <typename TListSample, typename TOutput, typename TCoordRep>
TOutput
ManifoldParzenWindowsListSampleFunction<TListSample, TOutput, TCoordRep>::Evaluate(
const InputMeasurementVectorType & measurement) const
{
try
{
unsigned int numberOfNeighbors =
std::min(this->m_EvaluationKNeighborhood, static_cast<unsigned int>(this->m_Gaussians.size()));
OutputType sum = 0.0;
if (numberOfNeighbors == this->m_Gaussians.size())
{
for (unsigned int j = 0; j < this->m_Gaussians.size(); j++)
{
sum += static_cast<OutputType>(this->m_Gaussians[j]->Evaluate(measurement));
}
}
else
{
typename TreeGeneratorType::KdTreeType::InstanceIdentifierVectorType neighbors;
this->m_KdTreeGenerator->GetOutput()->Search(measurement, numberOfNeighbors, neighbors);
for (unsigned int j = 0; j < numberOfNeighbors; j++)
{
sum += static_cast<OutputType>(this->m_Gaussians[neighbors[j]]->Evaluate(measurement));
}
}
return static_cast<OutputType>(sum / this->m_NormalizationFactor);
}
catch (...)
{
return 0;
}
}
/**
* Standard "PrintSelf" method
*/
template <typename TListSample, typename TOutput, typename TCoordRep>
void
ManifoldParzenWindowsListSampleFunction<TListSample, TOutput, TCoordRep>::PrintSelf(std::ostream & os,
Indent indent) const
{
os << indent << "Regularization sigma: " << this->m_RegularizationSigma << std::endl;
os << indent << "Evaluation K neighborhood: " << this->m_EvaluationKNeighborhood << std::endl;
if (this->m_CovarianceKNeighborhood > 0)
{
os << indent << "Covariance K neighborhood: " << this->m_CovarianceKNeighborhood << std::endl;
os << indent << "Kernel sigma: " << this->m_KernelSigma << std::endl;
}
}
} // end of namespace Statistics
} // end of namespace ants
} // end of namespace itk
#endif
|
