1 /*=========================================================================
2 Program: vv http://www.creatis.insa-lyon.fr/rio/vv
5 - University of LYON http://www.universite-lyon.fr/
6 - Léon Bérard cancer center http://oncora1.lyon.fnclcc.fr
7 - CREATIS CNRS laboratory http://www.creatis.insa-lyon.fr
9 This software is distributed WITHOUT ANY WARRANTY; without even
10 the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
11 PURPOSE. See the copyright notices for more information.
13 It is distributed under dual licence
15 - BSD See included LICENSE.txt file
16 - CeCILL-B http://www.cecill.info/licences/Licence_CeCILL-B_V1-en.html
17 ======================================================================-====*/
20 #include "clitkCommon.h"
24 #include "itkImageFileReader.h"
25 #include "itkImageSeriesReader.h"
26 #include "itkImageFileWriter.h"
27 #include "itkRecursiveGaussianImageFilter.h"
28 #include "itkResampleImageFilter.h"
29 #include "itkAffineTransform.h"
30 #include "itkNearestNeighborInterpolateImageFunction.h"
31 #include "itkLinearInterpolateImageFunction.h"
32 #include "itkBSplineInterpolateImageFunction.h"
33 #include "itkBSplineInterpolateImageFunctionWithLUT.h"
34 #include "itkCommand.h"
38 //--------------------------------------------------------------------
39 template <class TInputImage, class TOutputImage>
40 ResampleImageWithOptionsFilter<TInputImage, TOutputImage>::
41 ResampleImageWithOptionsFilter():itk::ImageToImageFilter<TInputImage, TOutputImage>() {
42 static const unsigned int dim = InputImageType::ImageDimension;
43 this->SetNumberOfRequiredInputs(1);
45 m_InterpolationType = NearestNeighbor;
46 m_GaussianFilteringEnabled = true;
48 m_BLUTSamplingFactor = 20;
49 m_LastDimensionIsTime = false;
50 m_Transform = TransformType::New();
51 if (dim == 4) m_LastDimensionIsTime = true; // by default 4D is 3D+t
52 for(unsigned int i=0; i<dim; i++) {
54 m_OutputSpacing[i] = -1;
55 m_GaussianSigma[i] = -1;
57 m_VerboseOptions = false;
59 //--------------------------------------------------------------------
62 //--------------------------------------------------------------------
63 template <class TInputImage, class TOutputImage>
65 ResampleImageWithOptionsFilter<TInputImage, TOutputImage>::
66 SetInput(const InputImageType * image) {
67 // Process object is not const-correct so the const casting is required.
68 this->SetNthInput(0, const_cast<InputImageType *>(image));
70 //--------------------------------------------------------------------
73 //--------------------------------------------------------------------
74 template <class TInputImage, class TOutputImage>
76 ResampleImageWithOptionsFilter<TInputImage, TOutputImage>::
77 GenerateInputRequestedRegion() {
78 // call the superclass's implementation of this method
79 Superclass::GenerateInputRequestedRegion();
81 // get pointers to the input and output
82 InputImagePointer inputPtr =
83 const_cast< TInputImage *>( this->GetInput() );
85 // Request the entire input image
86 InputImageRegionType inputRegion;
87 inputRegion = inputPtr->GetLargestPossibleRegion();
88 inputPtr->SetRequestedRegion(inputRegion);
90 //--------------------------------------------------------------------
93 //--------------------------------------------------------------------
94 template <class TInputImage, class TOutputImage>
96 ResampleImageWithOptionsFilter<TInputImage, TOutputImage>::
97 GenerateOutputInformation() {
98 static const unsigned int dim = InputImageType::ImageDimension;
101 if (!std::numeric_limits<InputImagePixelType>::is_signed) {
102 if ((m_InterpolationType == BSpline) ||
103 (m_InterpolationType == B_LUT)) {
104 std::cerr << "Warning : input pixel type is not signed, use bspline interpolation at your own risk ..." << std::endl;
109 InputImagePointer input = dynamic_cast<InputImageType*>(itk::ProcessObject::GetInput(0));
111 // Perform default implementation
112 Superclass::GenerateOutputInformation();
115 InputImageSpacingType inputSpacing = input->GetSpacing();
116 InputImageSizeType inputSize = input->GetLargestPossibleRegion().GetSize();
118 if (m_IsoSpacing != -1) { // apply isoSpacing
119 for(unsigned int i=0; i<dim; i++) {
120 m_OutputSpacing[i] = m_IsoSpacing;
121 m_OutputSize[i] = (int)lrint(inputSize[i]*inputSpacing[i]/m_OutputSpacing[i]);
125 if (m_OutputSpacing[0] != -1) { // apply spacing, compute size
126 for(unsigned int i=0; i<dim; i++) {
127 m_OutputSize[i] = (int)lrint(inputSize[i]*inputSpacing[i]/m_OutputSpacing[i]);
131 if (m_OutputSize[0] != 0) { // apply size, compute spacing
132 for(unsigned int i=0; i<dim; i++) {
133 m_OutputSpacing[i] = (double)inputSize[i]*inputSpacing[i]/(double)m_OutputSize[i];
136 else { // copy input size/spacing ... (no resampling)
137 m_OutputSize = inputSize;
138 m_OutputSpacing = inputSpacing;
143 // Special case for temporal image 2D+t or 3D+t
144 if (m_LastDimensionIsTime) {
146 m_OutputSize[l] = inputSize[l];
147 m_OutputSpacing[l] = inputSpacing[l];
151 OutputImagePointer outputImage = this->GetOutput(0);
152 OutputImageRegionType region;
153 region.SetSize(m_OutputSize);
154 region.SetIndex(input->GetLargestPossibleRegion().GetIndex());
155 outputImage->SetLargestPossibleRegion(region);
157 // Init Gaussian sigma
158 if (m_GaussianSigma[0] != -1) { // Gaussian filter set by user
159 m_GaussianFilteringEnabled = true;
162 if (m_GaussianFilteringEnabled) { // Automated sigma when downsample
163 for(unsigned int i=0; i<dim; i++) {
164 if (m_OutputSpacing[i] > inputSpacing[i]) { // downsample
165 m_GaussianSigma[i] = 0.5*m_OutputSpacing[i];// / inputSpacing[i]);
167 else m_GaussianSigma[i] = 0; // will be ignore after
171 if (m_GaussianFilteringEnabled && m_LastDimensionIsTime) {
172 m_GaussianSigma[dim-1] = 0;
175 //--------------------------------------------------------------------
178 //--------------------------------------------------------------------
179 template <class TInputImage, class TOutputImage>
181 ResampleImageWithOptionsFilter<TInputImage, TOutputImage>::
185 InputImagePointer input = dynamic_cast<InputImageType*>(itk::ProcessObject::GetInput(0));
186 static const unsigned int dim = InputImageType::ImageDimension;
188 // Create main Resample Image Filter
189 typedef itk::ResampleImageFilter<InputImageType,OutputImageType> FilterType;
190 typename FilterType::Pointer filter = FilterType::New();
191 filter->GraftOutput(this->GetOutput());
193 // Print options if needed
194 if (m_VerboseOptions) {
195 std::cout << "Output Spacing = " << m_OutputSpacing << std::endl
196 << "Output Size = " << m_OutputSize << std::endl
197 << "Gaussian = " << m_GaussianFilteringEnabled << std::endl;
198 if (m_GaussianFilteringEnabled)
199 std::cout << "Sigma = " << m_GaussianSigma << std::endl;
200 std::cout << "Interpol = ";
201 switch (m_InterpolationType) {
202 case NearestNeighbor: std::cout << "NearestNeighbor" << std::endl; break;
203 case Linear: std::cout << "Linear" << std::endl; break;
204 case BSpline: std::cout << "BSpline " << m_BSplineOrder << std::endl; break;
205 case B_LUT: std::cout << "B-LUT " << m_BSplineOrder << " " << m_BLUTSamplingFactor << std::endl; break;
207 std::cout << "Threads = " << this->GetNumberOfThreads() << std::endl;
208 std::cout << "LastDimIsTime = " << m_LastDimensionIsTime << std::endl;
211 // Instance of the transform object to be passed to the resample
212 // filter. By default, identity transform is applied
213 filter->SetTransform(m_Transform);
214 filter->SetSize(m_OutputSize);
215 filter->SetOutputSpacing(m_OutputSpacing);
216 filter->SetOutputOrigin(input->GetOrigin());
217 filter->SetDefaultPixelValue(m_DefaultPixelValue);
218 filter->SetNumberOfThreads(this->GetNumberOfThreads());
220 // Select interpolator
221 switch (m_InterpolationType) {
222 case NearestNeighbor:
224 typedef itk::NearestNeighborInterpolateImageFunction<InputImageType, double> InterpolatorType;
225 typename InterpolatorType::Pointer interpolator = InterpolatorType::New();
226 filter->SetInterpolator(interpolator);
231 typedef itk::LinearInterpolateImageFunction<InputImageType, double> InterpolatorType;
232 typename InterpolatorType::Pointer interpolator = InterpolatorType::New();
233 filter->SetInterpolator(interpolator);
238 typedef itk::BSplineInterpolateImageFunction<InputImageType, double> InterpolatorType;
239 typename InterpolatorType::Pointer interpolator = InterpolatorType::New();
240 interpolator->SetSplineOrder(m_BSplineOrder);
241 filter->SetInterpolator(interpolator);
246 typedef itk::BSplineInterpolateImageFunctionWithLUT<InputImageType, double> InterpolatorType;
247 typename InterpolatorType::Pointer interpolator = InterpolatorType::New();
248 interpolator->SetSplineOrder(m_BSplineOrder);
249 interpolator->SetLUTSamplingFactor(m_BLUTSamplingFactor);
250 filter->SetInterpolator(interpolator);
255 // Initial Gaussian blurring if needed
256 typedef itk::RecursiveGaussianImageFilter<InputImageType, InputImageType> GaussianFilterType;
257 std::vector<typename GaussianFilterType::Pointer> gaussianFilters;
258 if (m_GaussianFilteringEnabled) {
259 for(unsigned int i=0; i<dim; i++) {
260 if (m_GaussianSigma[i] != 0) {
261 gaussianFilters.push_back(GaussianFilterType::New());
262 gaussianFilters[i]->SetDirection(i);
263 gaussianFilters[i]->SetOrder(GaussianFilterType::ZeroOrder);
264 gaussianFilters[i]->SetNormalizeAcrossScale(false);
265 gaussianFilters[i]->SetSigma(m_GaussianSigma[i]); // in millimeter !
266 if (gaussianFilters.size() == 1) { // first
267 gaussianFilters[0]->SetInput(input);
270 gaussianFilters[i]->SetInput(gaussianFilters[i-1]->GetOutput());
274 if (gaussianFilters.size() > 0) {
275 filter->SetInput(gaussianFilters[gaussianFilters.size()-1]->GetOutput());
277 else filter->SetInput(input);
279 else filter->SetInput(input);
285 this->GraftOutput(filter->GetOutput());
287 //--------------------------------------------------------------------