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
|
/*=========================================================================
*
* 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 itkConvolutionImageFilter_h
#define itkConvolutionImageFilter_h
#include "itkConvolutionImageFilterBase.h"
#include "itkProgressAccumulator.h"
#include "itkZeroFluxNeumannBoundaryCondition.h"
namespace itk
{
/**
* \class ConvolutionImageFilter
* \brief Convolve a given image with an arbitrary image kernel.
*
* This filter operates by centering the flipped kernel at each pixel
* in the image and computing the inner product between pixel values
* in the image and pixel values in the kernel. The center of the
* kernel is defined as \f$ \lfloor (2*i+s-1)/2 \rfloor \f$ where
* \f$i\f$ is the index and \f$s\f$ is the size of the largest
* possible region of the kernel image. For kernels with odd sizes in
* all dimensions, this corresponds to the center pixel. If a
* dimension of the kernel image has an even size, then the center
* index of the kernel in that dimension will be the largest integral
* index that is less than the continuous index of the image center.
*
* The kernel can optionally be normalized to sum to 1 using
* NormalizeOn(). Normalization is off by default.
*
* \warning This filter ignores the spacing, origin, and orientation
* of the kernel image and treats them as identical to those in the
* input image.
*
* This code was contributed in the Insight Journal paper:
*
* "Image Kernel Convolution"
* by Tustison N., Gee J.
* https://www.insight-journal.org/browse/publication/208
*
* \author Nicholas J. Tustison
* \author James C. Gee
* \ingroup ITKConvolution
*
* \sphinx
* \sphinxexample{Filtering/Convolution/ConvolveImageWithKernel,Convolve Image With Kernel}
* \endsphinx
*/
template <typename TInputImage, typename TKernelImage = TInputImage, typename TOutputImage = TInputImage>
class ITK_TEMPLATE_EXPORT ConvolutionImageFilter
: public ConvolutionImageFilterBase<TInputImage, TKernelImage, TOutputImage>
{
public:
ITK_DISALLOW_COPY_AND_MOVE(ConvolutionImageFilter);
using Self = ConvolutionImageFilter;
using Superclass = ConvolutionImageFilterBase<TInputImage, TOutputImage>;
using Pointer = SmartPointer<Self>;
using ConstPointer = SmartPointer<const Self>;
/** Method for creation through the object factory. */
itkNewMacro(Self);
/** \see LightObject::GetNameOfClass() */
itkOverrideGetNameOfClassMacro(ConvolutionImageFilter);
/** Dimensionality of input and output data is assumed to be the same. */
static constexpr unsigned int ImageDimension = TInputImage::ImageDimension;
using InputImageType = TInputImage;
using OutputImageType = TOutputImage;
using KernelImageType = TKernelImage;
using InputPixelType = typename InputImageType::PixelType;
using OutputPixelType = typename OutputImageType::PixelType;
using KernelPixelType = typename KernelImageType::PixelType;
using InputIndexType = typename InputImageType::IndexType;
using OutputIndexType = typename OutputImageType::IndexType;
using KernelIndexType = typename KernelImageType::IndexType;
using InputSizeType = typename InputImageType::SizeType;
using OutputSizeType = typename OutputImageType::SizeType;
using KernelSizeType = typename KernelImageType::SizeType;
using InputRegionType = typename InputImageType::RegionType;
using OutputRegionType = typename OutputImageType::RegionType;
using KernelRegionType = typename KernelImageType::RegionType;
protected:
ConvolutionImageFilter() = default;
~ConvolutionImageFilter() override = default;
/** ConvolutionImageFilter needs the entire image kernel, which in
* general is going to be a different size then the output requested
* region. As such, this filter needs to provide an implementation
* for GenerateInputRequestedRegion() in order to inform the
* pipeline execution model.
*
* \sa ProcessObject::GenerateInputRequestedRegion() */
void
GenerateInputRequestedRegion() override;
/** This filter uses a minipipeline to compute the output. */
void
GenerateData() override;
/** The kernel needs padding if any of the sizes of its dimensions is
* even. This method checks for this condition. */
bool
GetKernelNeedsPadding() const;
/** Calculates the padding width needed to make each dimension odd. */
KernelSizeType
GetKernelPadSize() const;
/** Calculates the radius of the kernel. */
template <typename TImage>
KernelSizeType
GetKernelRadius(const TImage * kernelImage) const;
private:
template <typename TImage>
void
ComputeConvolution(const TImage * kernelImage, ProgressAccumulator * progress);
};
} // namespace itk
#ifndef ITK_MANUAL_INSTANTIATION
# include "itkConvolutionImageFilter.hxx"
#endif
#endif
|
