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
|
/*=========================================================================
*
* Copyright NumFOCUS
*
* 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
*
* https://www.apache.org/licenses/LICENSE-2.0.txt
*
* 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 itkHessianToObjectnessMeasureImageFilter_h
#define itkHessianToObjectnessMeasureImageFilter_h
#include "itkSymmetricSecondRankTensor.h"
#include "itkImageToImageFilter.h"
namespace itk
{
/**
* \class HessianToObjectnessMeasureImageFilter
* \brief A filter to enhance M-dimensional objects in N-dimensional images
*
* The objectness measure is a generalization of Frangi's vesselness measure,
* which is based on the analysis of the Hessian eigen system. The filter
* can enhance blob-like structures (M=0), vessel-like structures (M=1), 2D
* plate-like structures (M=2), hyper-plate-like structures (M=3) in N-dimensional
* images, with M<N.
* The filter takes an image of a Hessian pixels ( SymmetricSecondRankTensor pixels
* pixels ) and produces an enhanced image. The Hessian input image can be produced
* using itk::HessianRecursiveGaussianImageFilter.
*
*
* \par References
* Frangi, AF, Niessen, WJ, Vincken, KL, & Viergever, MA (1998). Multiscale Vessel
* Enhancement Filtering. In Wells, WM, Colchester, A, & Delp, S, Editors, MICCAI '98
* Medical Image Computing and Computer-Assisted Intervention, Lecture Notes in Computer
* Science, pages 130-137, Springer Verlag, 1998.
*
* Additional information can be from in the Insight Journal:
* https://www.insight-journal.org/browse/publication/175
*
* \author Luca Antiga Ph.D. Medical Imaging Unit,
* Bioengineering Department, Mario Negri Institute, Italy.
*
* \sa MultiScaleHessianBasedMeasureImageFilter
* \sa Hessian3DToVesselnessMeasureImageFilter
* \sa HessianRecursiveGaussianImageFilter
* \sa SymmetricEigenAnalysisImageFilter
* \sa SymmetricSecondRankTensor
*
* \ingroup ITKImageFeature
*/
template <typename TInputImage, typename TOutputImage>
class ITK_TEMPLATE_EXPORT HessianToObjectnessMeasureImageFilter : public ImageToImageFilter<TInputImage, TOutputImage>
{
public:
ITK_DISALLOW_COPY_AND_MOVE(HessianToObjectnessMeasureImageFilter);
/** Standard class type aliases. */
using Self = HessianToObjectnessMeasureImageFilter;
using Superclass = ImageToImageFilter<TInputImage, TOutputImage>;
using Pointer = SmartPointer<Self>;
using ConstPointer = SmartPointer<const Self>;
using typename Superclass::InputImageType;
using typename Superclass::OutputImageType;
using InputPixelType = typename InputImageType::PixelType;
using OutputPixelType = typename OutputImageType::PixelType;
using OutputImageRegionType = typename OutputImageType::RegionType;
/** Image dimension */
static constexpr unsigned int ImageDimension = InputImageType::ImageDimension;
using EigenValueType = double;
using EigenValueArrayType = itk::FixedArray<EigenValueType, Self::ImageDimension>;
/** Method for creation through the object factory. */
itkNewMacro(Self);
/** \see LightObject::GetNameOfClass() */
itkOverrideGetNameOfClassMacro(HessianToObjectnessMeasureImageFilter);
/** Set/Get Alpha, the weight corresponding to R_A
* (the ratio of the smallest eigenvalue that has to be large to the larger ones).
* Smaller values lead to increased sensitivity to the object dimensionality. */
itkSetMacro(Alpha, double);
itkGetConstMacro(Alpha, double);
/** Set/Get Beta, the weight corresponding to R_B
* (the ratio of the largest eigenvalue that has to be small to the larger ones).
* Smaller values lead to increased sensitivity to the object dimensionality. */
itkSetMacro(Beta, double);
itkGetConstMacro(Beta, double);
/** Set/Get Gamma, the weight corresponding to S
* (the Frobenius norm of the Hessian matrix, or second-order structureness) */
itkSetMacro(Gamma, double);
itkGetConstMacro(Gamma, double);
/** Toggle scaling the objectness measure with the magnitude of the largest
absolute eigenvalue */
itkSetMacro(ScaleObjectnessMeasure, bool);
itkGetConstMacro(ScaleObjectnessMeasure, bool);
itkBooleanMacro(ScaleObjectnessMeasure);
/** Set/Get the dimensionality of the object (0: points (blobs),
* 1: lines (vessels), 2: planes (plate-like structures), 3: hyper-planes.
* ObjectDimension must be smaller than ImageDimension. */
itkSetMacro(ObjectDimension, unsigned int);
itkGetConstMacro(ObjectDimension, unsigned int);
/** Enhance bright structures on a dark background if true, the opposite if
false. Default is "On" (equivalent to vesselness). */
itkSetMacro(BrightObject, bool);
itkGetConstMacro(BrightObject, bool);
itkBooleanMacro(BrightObject);
#ifdef ITK_USE_CONCEPT_CHECKING
// Begin concept checking
itkConceptMacro(DoubleConvertibleToOutputCheck, (Concept::Convertible<double, OutputPixelType>));
// End concept checking
#endif
protected:
HessianToObjectnessMeasureImageFilter();
~HessianToObjectnessMeasureImageFilter() override = default;
void
PrintSelf(std::ostream & os, Indent indent) const override;
void
VerifyPreconditions() ITKv5_CONST override;
void
DynamicThreadedGenerateData(const OutputImageRegionType & outputRegionForThread) override;
private:
// functor used to sort the eigenvalues are to be sorted
// |e1|<=|e2|<=...<=|eN|
//
// Returns ( itk::Math::abs(a) < itk::Math::abs(b) )
struct AbsLessCompare
{
bool
operator()(EigenValueType a, EigenValueType b)
{
return itk::Math::abs(a) < itk::Math::abs(b);
}
};
double m_Alpha{ 0.5 };
double m_Beta{ 0.5 };
double m_Gamma{ 5.0 };
unsigned int m_ObjectDimension{ 1 };
bool m_BrightObject{ true };
bool m_ScaleObjectnessMeasure{ true };
};
} // end namespace itk
#ifndef ITK_MANUAL_INSTANTIATION
# include "itkHessianToObjectnessMeasureImageFilter.hxx"
#endif
#endif
|
