| 1 |
|
/*========================================================================= |
| 2 |
|
|
| 3 |
|
Program: Insight Segmentation & Registration Toolkit |
| 4 |
|
Module: $RCSfile: itkGaussianBlurImageFunction.txx.html,v $ |
| 5 |
|
Language: C++ |
| 6 |
|
Date: $Date: 2006/01/17 19:15:36 $ |
| 7 |
|
Version: $Revision: 1.4 $ |
| 8 |
|
|
| 9 |
|
Copyright (c) Insight Software Consortium. All rights reserved. |
| 10 |
|
See ITKCopyright.txt or http://www.itk.org/HTML/Copyright.htm for details. |
| 11 |
|
|
| 12 |
|
This software is distributed WITHOUT ANY WARRANTY; without even |
| 13 |
|
the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR |
| 14 |
|
PURPOSE. See the above copyright notices for more information. |
| 15 |
|
|
| 16 |
|
=========================================================================*/ |
| 17 |
DEF |
#ifndef _itkGaussianBlurImageFunction_txx |
| 18 |
DEF |
#define _itkGaussianBlurImageFunction_txx |
| 19 |
|
|
| 20 |
|
#include "itkGaussianBlurImageFunction.h" |
| 21 |
|
#include "itkImageLinearIteratorWithIndex.h" |
| 22 |
|
|
| 23 |
|
namespace itk |
| 24 |
|
{ |
| 25 |
|
|
| 26 |
|
/** Set the Input Image */ |
| 27 |
|
template <class TInputImage,class TOutput> |
| 28 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 29 |
|
::GaussianBlurImageFunction() |
| 30 |
|
{ |
| 31 |
|
typename GaussianFunctionType::ArrayType mean; |
| 32 |
|
mean[0]=0.0f; |
| 33 |
|
for(unsigned int i=0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 34 |
|
{ |
| 35 |
|
m_Sigma[i] = 1.0f; |
| 36 |
|
m_MaximumError[i] = 0.001f; |
| 37 |
|
m_MaximumKernelWidth = 32; |
| 38 |
|
m_Extent[i] = 1.0f; |
| 39 |
|
} |
| 40 |
|
m_UseImageSpacing = true; |
| 41 |
|
|
| 42 |
|
m_GaussianFunction = GaussianFunctionType::New(); |
| 43 |
|
m_GaussianFunction->SetMean(mean); |
| 44 |
|
m_GaussianFunction->SetNormalized(false); // faster |
| 45 |
|
m_OperatorImageFunction = OperatorImageFunctionType::New(); |
| 46 |
|
m_InternalImage = InternalImageType::New(); |
| 47 |
|
this->RecomputeGaussianKernel(); |
| 48 |
|
m_Caster = CastImageFilterType::New(); |
| 49 |
|
} |
| 50 |
|
|
| 51 |
|
/** Set the input image */ |
| 52 |
|
template <class TInputImage,class TOutput> |
| 53 |
|
void |
| 54 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 55 |
|
::SetInputImage( const InputImageType * ptr ) |
| 56 |
|
{ |
| 57 |
|
Superclass::SetInputImage(ptr); |
| 58 |
|
m_Caster->SetInput(ptr); |
| 59 |
|
m_Caster->Update(); |
| 60 |
|
m_OperatorImageFunction->SetInputImage(m_Caster->GetOutput()); |
| 61 |
|
} |
| 62 |
|
|
| 63 |
|
|
| 64 |
|
/** Print self method */ |
| 65 |
|
template <class TInputImage,class TOutput> |
| 66 |
|
void |
| 67 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 68 |
|
::PrintSelf(std::ostream& os, Indent indent) const |
| 69 |
|
{ |
| 70 |
|
this->Superclass::PrintSelf(os,indent); |
| 71 |
|
|
| 72 |
|
for(unsigned int i=0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 73 |
|
{ |
| 74 |
|
os << indent << "Sigma["<< i << "] : " << m_Sigma[i] << std::endl; |
| 75 |
LEN |
os << indent << "MaximumError["<< i << "] : " << m_MaximumError[i] << std::endl; |
| 76 |
|
os << indent << "Extent["<< i << "] : " << m_Extent[i] << std::endl; |
| 77 |
|
} |
| 78 |
|
os << indent << "MaximumKernelWidth: " << m_MaximumKernelWidth << std::endl; |
| 79 |
|
os << indent << "UseImageSpacing: " << m_UseImageSpacing << std::endl; |
| 80 |
|
|
| 81 |
|
os << indent << "Internal Image : " << m_InternalImage << std::endl; |
| 82 |
|
os << indent << "Image Caster : " << m_Caster << std::endl; |
| 83 |
|
|
| 84 |
|
} |
| 85 |
|
|
| 86 |
|
/** Set the variance of the gaussian in each direction */ |
| 87 |
|
template <class TInputImage,class TOutput> |
| 88 |
|
void |
| 89 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 90 |
|
::SetSigma(const double* sigma) |
| 91 |
|
{ |
| 92 |
|
unsigned int i; |
| 93 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 94 |
|
{ |
| 95 |
|
if ( sigma[i] != m_Sigma[i] ) |
| 96 |
|
{ |
| 97 |
|
break; |
| 98 |
|
} |
| 99 |
|
} |
| 100 |
|
if ( i < itkGetStaticConstMacro(ImageDimension) ) |
| 101 |
|
{ |
| 102 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 103 |
|
{ |
| 104 |
|
m_Sigma[i] = sigma[i]; |
| 105 |
|
} |
| 106 |
|
this->RecomputeGaussianKernel(); |
| 107 |
|
} |
| 108 |
|
} |
| 109 |
|
|
| 110 |
|
|
| 111 |
|
/** Set the variance of the gaussian in each direction */ |
| 112 |
|
template <class TInputImage,class TOutput> |
| 113 |
|
void |
| 114 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 115 |
|
::SetSigma( const double sigma) |
| 116 |
|
{ |
| 117 |
|
unsigned int i; |
| 118 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 119 |
|
{ |
| 120 |
|
if ( sigma != m_Sigma[i] ) |
| 121 |
|
{ |
| 122 |
|
break; |
| 123 |
|
} |
| 124 |
|
} |
| 125 |
|
if ( i < itkGetStaticConstMacro(ImageDimension) ) |
| 126 |
|
{ |
| 127 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 128 |
|
{ |
| 129 |
|
m_Sigma[i] = sigma; |
| 130 |
|
} |
| 131 |
|
this->RecomputeGaussianKernel(); |
| 132 |
|
} |
| 133 |
|
} |
| 134 |
|
|
| 135 |
|
/** Set the extent of the gaussian in each direction */ |
| 136 |
|
template <class TInputImage,class TOutput> |
| 137 |
|
void |
| 138 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 139 |
|
::SetExtent(const double* extent) |
| 140 |
|
{ |
| 141 |
|
unsigned int i; |
| 142 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 143 |
|
{ |
| 144 |
|
if ( extent[i] != m_Extent[i] ) |
| 145 |
|
{ |
| 146 |
|
break; |
| 147 |
|
} |
| 148 |
|
} |
| 149 |
|
if ( i < itkGetStaticConstMacro(ImageDimension) ) |
| 150 |
|
{ |
| 151 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 152 |
|
{ |
| 153 |
|
m_Extent[i] = extent[i]; |
| 154 |
|
} |
| 155 |
|
this->RecomputeGaussianKernel(); |
| 156 |
|
} |
| 157 |
|
} |
| 158 |
|
|
| 159 |
|
|
| 160 |
|
/** Set the extent of the gaussian in each direction */ |
| 161 |
|
template <class TInputImage,class TOutput> |
| 162 |
|
void |
| 163 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 164 |
|
::SetExtent( const double extent) |
| 165 |
|
{ |
| 166 |
|
unsigned int i; |
| 167 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 168 |
|
{ |
| 169 |
|
if ( extent != m_Extent[i] ) |
| 170 |
|
{ |
| 171 |
|
break; |
| 172 |
|
} |
| 173 |
|
} |
| 174 |
|
if ( i < itkGetStaticConstMacro(ImageDimension) ) |
| 175 |
|
{ |
| 176 |
|
for (i=0; i<itkGetStaticConstMacro(ImageDimension); i++) |
| 177 |
|
{ |
| 178 |
|
m_Extent[i] = extent; |
| 179 |
|
} |
| 180 |
|
this->RecomputeGaussianKernel(); |
| 181 |
|
} |
| 182 |
|
} |
| 183 |
|
|
| 184 |
|
/** Recompute the gaussian kernel used to evaluate indexes |
| 185 |
|
* And allocate the internal image for processing depending on |
| 186 |
|
* the size of the operator */ |
| 187 |
|
template <class TInputImage,class TOutput> |
| 188 |
|
void |
| 189 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 190 |
|
::RecomputeGaussianKernel() |
| 191 |
|
{ |
| 192 |
|
|
| 193 |
|
typename InternalImageType::SizeType size; |
| 194 |
|
// Compute the convolution of each kernel in each direction |
| 195 |
LEN |
for(unsigned int direction=0;direction<itkGetStaticConstMacro(ImageDimension);direction++) |
| 196 |
|
{ |
| 197 |
|
GaussianOperatorType gaussianOperator; |
| 198 |
|
|
| 199 |
|
gaussianOperator.SetDirection(direction); |
| 200 |
|
gaussianOperator.SetMaximumError(m_MaximumError[direction]); |
| 201 |
|
gaussianOperator.SetMaximumKernelWidth(m_MaximumKernelWidth); |
| 202 |
|
|
| 203 |
|
if( (m_UseImageSpacing == true) && (this->GetInputImage()) ) |
| 204 |
|
{ |
| 205 |
|
if (this->GetInputImage()->GetSpacing()[direction] == 0.0) |
| 206 |
|
{ |
| 207 |
|
itkExceptionMacro(<< "Pixel spacing cannot be zero"); |
| 208 |
|
} |
| 209 |
|
else |
| 210 |
|
{ |
| 211 |
LEN |
gaussianOperator.SetVariance(m_Sigma[direction]*m_Sigma[direction] / this->GetInputImage()->GetSpacing()[direction]); |
| 212 |
|
} |
| 213 |
|
} |
| 214 |
|
else |
| 215 |
|
{ |
| 216 |
|
gaussianOperator.SetVariance(m_Sigma[direction]*m_Sigma[direction]); |
| 217 |
|
} |
| 218 |
|
|
| 219 |
|
gaussianOperator.CreateDirectional(); |
| 220 |
|
m_OperatorArray[direction] = gaussianOperator; |
| 221 |
|
size[direction] = gaussianOperator.GetSize()[direction]; |
| 222 |
|
} |
| 223 |
|
|
| 224 |
|
// Allocate the internal image |
| 225 |
|
m_InternalImage = InternalImageType::New(); |
| 226 |
|
typename InternalImageType::RegionType region; |
| 227 |
|
region.SetSize(size); |
| 228 |
|
m_InternalImage->SetRegions(region); |
| 229 |
|
m_InternalImage->Allocate(); |
| 230 |
|
m_InternalImage->FillBuffer(0); |
| 231 |
|
} |
| 232 |
|
|
| 233 |
|
/** Evaluate the function at the specifed point */ |
| 234 |
|
template <class TInputImage,class TOutput> |
| 235 |
|
TOutput |
| 236 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 237 |
|
::EvaluateAtIndex(const IndexType& index) const |
| 238 |
|
{ |
| 239 |
|
// First time we use the complete image and fill the internal image |
| 240 |
|
m_OperatorImageFunction->SetInputImage(m_Caster->GetOutput()); |
| 241 |
|
m_OperatorImageFunction->SetOperator(m_OperatorArray[0]); |
| 242 |
|
|
| 243 |
|
// if 1D Image we return the result |
| 244 |
|
if(itkGetStaticConstMacro(ImageDimension) == 1) |
| 245 |
|
{ |
| 246 |
|
return m_OperatorImageFunction->EvaluateAtIndex(index); |
| 247 |
|
} |
| 248 |
|
|
| 249 |
|
// Compute the centered index fo the neighborhood |
| 250 |
|
IndexType centerIndex; |
| 251 |
|
for(unsigned int i=0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 252 |
|
{ |
| 253 |
LEN |
centerIndex[i] = (unsigned long)((float)m_InternalImage->GetBufferedRegion().GetSize()[i]/2.0); |
| 254 |
|
} |
| 255 |
|
|
| 256 |
|
// first direction |
| 257 |
|
typename InternalImageType::IndexType ind; |
| 258 |
|
ind = index; |
| 259 |
|
|
| 260 |
|
//Define the region of the iterator |
| 261 |
|
typename InternalImageType::RegionType region; |
| 262 |
LEN |
typename InternalImageType::SizeType size = m_InternalImage->GetBufferedRegion().GetSize(); |
| 263 |
|
size[0]=1; |
| 264 |
|
region.SetSize(size); |
| 265 |
|
|
| 266 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 267 |
|
{ |
| 268 |
|
if(i != 0) |
| 269 |
|
{ |
| 270 |
|
ind[i] -= centerIndex[i]; |
| 271 |
|
} |
| 272 |
|
} |
| 273 |
|
region.SetIndex(ind); |
| 274 |
|
|
| 275 |
|
typename InternalImageType::RegionType regionN; |
| 276 |
|
regionN.SetSize(size); |
| 277 |
|
ind = centerIndex; |
| 278 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 279 |
|
{ |
| 280 |
|
if(i != 0) |
| 281 |
|
{ |
| 282 |
|
ind[i] = 0; |
| 283 |
|
} |
| 284 |
|
} |
| 285 |
|
regionN.SetIndex(ind); |
| 286 |
|
|
| 287 |
LEN |
itk::ImageLinearConstIteratorWithIndex<InternalImageType> it(m_Caster->GetOutput(),region); |
| 288 |
LEN |
itk::ImageLinearIteratorWithIndex<InternalImageType> itN(m_InternalImage,regionN); |
| 289 |
|
it.SetDirection(1); |
| 290 |
|
itN.SetDirection(1); |
| 291 |
|
it.GoToBeginOfLine(); |
| 292 |
|
itN.GoToBeginOfLine(); |
| 293 |
|
while(!it.IsAtEnd()) |
| 294 |
|
{ |
| 295 |
|
while(!it.IsAtEndOfLine()) |
| 296 |
|
{ |
| 297 |
LEN |
if(m_Caster->GetOutput()->GetLargestPossibleRegion().IsInside(it.GetIndex())) |
| 298 |
|
{ |
| 299 |
|
itN.Set(m_OperatorImageFunction->EvaluateAtIndex(it.GetIndex())); |
| 300 |
|
} |
| 301 |
|
++it; |
| 302 |
|
++itN; |
| 303 |
|
} |
| 304 |
|
it.NextLine(); |
| 305 |
|
itN.NextLine(); |
| 306 |
|
} |
| 307 |
|
|
| 308 |
|
// Do the convolution in other directions |
| 309 |
LEN |
for(unsigned int direction=1;direction<itkGetStaticConstMacro(ImageDimension);direction++) |
| 310 |
|
{ |
| 311 |
|
|
| 312 |
|
size[direction] = 1; |
| 313 |
|
ind = centerIndex; |
| 314 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 315 |
|
{ |
| 316 |
|
if(i > direction) |
| 317 |
|
{ |
| 318 |
|
ind[i] = 0; |
| 319 |
|
} |
| 320 |
|
} |
| 321 |
|
region.SetSize(size); |
| 322 |
|
region.SetIndex(ind); |
| 323 |
|
|
| 324 |
|
|
| 325 |
|
m_OperatorImageFunction->SetInputImage(m_InternalImage); |
| 326 |
|
m_OperatorImageFunction->SetOperator(m_OperatorArray[direction]); |
| 327 |
|
|
| 328 |
LEN |
itk::ImageLinearIteratorWithIndex<InternalImageType> it(m_InternalImage,region); |
| 329 |
|
|
| 330 |
|
unsigned int dir = direction +1; |
| 331 |
|
if(dir == itkGetStaticConstMacro(ImageDimension)) |
| 332 |
|
{ |
| 333 |
|
dir = itkGetStaticConstMacro(ImageDimension)-1; |
| 334 |
|
} |
| 335 |
|
|
| 336 |
|
it.SetDirection(dir); |
| 337 |
|
it.GoToBeginOfLine(); |
| 338 |
|
while(!it.IsAtEnd()) |
| 339 |
|
{ |
| 340 |
|
while(!it.IsAtEndOfLine()) |
| 341 |
|
{ |
| 342 |
|
it.Set(m_OperatorImageFunction->EvaluateAtIndex(it.GetIndex())); |
| 343 |
|
++it; |
| 344 |
|
} |
| 345 |
|
it.NextLine(); |
| 346 |
|
} |
| 347 |
|
} |
| 348 |
|
|
| 349 |
|
return m_InternalImage->GetPixel(centerIndex); |
| 350 |
|
|
| 351 |
|
} |
| 352 |
|
|
| 353 |
|
|
| 354 |
|
/** Recompute the gaussian kernel used to evaluate indexes |
| 355 |
|
* The variance should be uniform */ |
| 356 |
|
template <class TInputImage,class TOutput> |
| 357 |
|
void |
| 358 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 359 |
|
::RecomputeContinuousGaussianKernel(const double* offset) const |
| 360 |
|
{ |
| 361 |
LEN |
for(unsigned int direction=0;direction<itkGetStaticConstMacro(ImageDimension);direction++) |
| 362 |
|
{ |
| 363 |
|
NeighborhoodType gaussianNeighborhood; |
| 364 |
|
typename GaussianFunctionType::InputType pt; |
| 365 |
|
typename NeighborhoodType::SizeType size; |
| 366 |
|
size.Fill(0); |
| 367 |
|
size[direction] = (unsigned long)(m_Sigma[direction]*m_Extent[direction]); |
| 368 |
|
|
| 369 |
|
gaussianNeighborhood.SetRadius(size); |
| 370 |
|
|
| 371 |
|
typename NeighborhoodType::Iterator it = gaussianNeighborhood.Begin(); |
| 372 |
|
|
| 373 |
|
itk::FixedArray<double,1> s; |
| 374 |
|
s[0]=m_Sigma[direction]; |
| 375 |
|
m_GaussianFunction->SetSigma(s); |
| 376 |
|
|
| 377 |
|
unsigned int i=0; |
| 378 |
|
float sum = 0; |
| 379 |
|
while(it != gaussianNeighborhood.End() ) |
| 380 |
|
{ |
| 381 |
|
pt[0]= gaussianNeighborhood.GetOffset(i)[direction]-offset[direction]; |
| 382 |
|
if( (m_UseImageSpacing == true) && (this->GetInputImage()) ) |
| 383 |
|
{ |
| 384 |
|
if (this->GetInputImage()->GetSpacing()[direction] == 0.0) |
| 385 |
|
{ |
| 386 |
|
itkExceptionMacro(<< "Pixel spacing cannot be zero"); |
| 387 |
|
} |
| 388 |
|
else |
| 389 |
|
{ |
| 390 |
|
pt[0] *= this->GetInputImage()->GetSpacing()[direction]; |
| 391 |
|
} |
| 392 |
|
} |
| 393 |
|
|
| 394 |
|
(*it)= m_GaussianFunction->Evaluate(pt); |
| 395 |
|
sum += (*it); |
| 396 |
|
i++; |
| 397 |
|
it++; |
| 398 |
|
} |
| 399 |
|
|
| 400 |
|
// Make the filter DC-Constant |
| 401 |
|
it = gaussianNeighborhood.Begin(); |
| 402 |
|
while(it != gaussianNeighborhood.End() ) |
| 403 |
|
{ |
| 404 |
|
(*it) /= sum; |
| 405 |
|
it++; |
| 406 |
|
} |
| 407 |
|
m_ContinuousOperatorArray[direction] = gaussianNeighborhood; |
| 408 |
|
} |
| 409 |
|
} |
| 410 |
|
|
| 411 |
|
/** Evaluate the function at the specifed point */ |
| 412 |
|
template <class TInputImage,class TOutput> |
| 413 |
|
TOutput |
| 414 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 415 |
|
::Evaluate(const PointType& point) const |
| 416 |
|
{ |
| 417 |
|
|
| 418 |
|
IndexType index; |
| 419 |
|
|
| 420 |
|
double offset[itkGetStaticConstMacro(ImageDimension)]; |
| 421 |
|
for(unsigned int i=0; i<itkGetStaticConstMacro(ImageDimension);i++) |
| 422 |
|
{ |
| 423 |
|
index[i] = (unsigned long)point[i]; |
| 424 |
|
offset[i] = point[i]-index[i]; |
| 425 |
|
} |
| 426 |
|
|
| 427 |
|
this->RecomputeContinuousGaussianKernel(offset); |
| 428 |
|
|
| 429 |
|
// First time we use the complete image and fill the internal image |
| 430 |
|
m_OperatorImageFunction->SetInputImage(m_Caster->GetOutput()); |
| 431 |
|
m_OperatorImageFunction->SetOperator(m_ContinuousOperatorArray[0]); |
| 432 |
|
|
| 433 |
|
// if 1D Image we return the result |
| 434 |
|
if(itkGetStaticConstMacro(ImageDimension) == 1) |
| 435 |
|
{ |
| 436 |
|
return m_OperatorImageFunction->EvaluateAtIndex(index); |
| 437 |
|
} |
| 438 |
|
|
| 439 |
|
// Compute the centered index fo the neighborhood |
| 440 |
|
IndexType centerIndex; |
| 441 |
|
for(unsigned int i=0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 442 |
|
{ |
| 443 |
LEN |
centerIndex[i] = (unsigned long)((float)m_InternalImage->GetBufferedRegion().GetSize()[i]/2.0); |
| 444 |
|
} |
| 445 |
|
|
| 446 |
|
// first direction |
| 447 |
|
typename InternalImageType::IndexType ind; |
| 448 |
|
ind = index; |
| 449 |
|
|
| 450 |
|
//Define the region of the iterator |
| 451 |
|
typename InternalImageType::RegionType region; |
| 452 |
LEN |
typename InternalImageType::SizeType size = m_InternalImage->GetBufferedRegion().GetSize(); |
| 453 |
|
size[0]=1; |
| 454 |
|
region.SetSize(size); |
| 455 |
|
|
| 456 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 457 |
|
{ |
| 458 |
|
if(i != 0) |
| 459 |
|
{ |
| 460 |
|
ind[i] -= centerIndex[i]; |
| 461 |
|
} |
| 462 |
|
} |
| 463 |
|
region.SetIndex(ind); |
| 464 |
|
|
| 465 |
|
typename InternalImageType::RegionType regionN; |
| 466 |
|
regionN.SetSize(size); |
| 467 |
|
ind = centerIndex; |
| 468 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 469 |
|
{ |
| 470 |
|
if(i != 0) |
| 471 |
|
{ |
| 472 |
|
ind[i] = 0; |
| 473 |
|
} |
| 474 |
|
} |
| 475 |
|
regionN.SetIndex(ind); |
| 476 |
|
|
| 477 |
LEN |
itk::ImageLinearConstIteratorWithIndex<InternalImageType> it(m_Caster->GetOutput(),region); |
| 478 |
LEN |
itk::ImageLinearIteratorWithIndex<InternalImageType> itN(m_InternalImage,regionN); |
| 479 |
|
it.SetDirection(1); |
| 480 |
|
itN.SetDirection(1); |
| 481 |
|
it.GoToBeginOfLine(); |
| 482 |
|
itN.GoToBeginOfLine(); |
| 483 |
|
while(!it.IsAtEnd()) |
| 484 |
|
{ |
| 485 |
|
while(!it.IsAtEndOfLine()) |
| 486 |
|
{ |
| 487 |
LEN |
if(m_Caster->GetOutput()->GetLargestPossibleRegion().IsInside(it.GetIndex())) |
| 488 |
|
{ |
| 489 |
|
itN.Set(m_OperatorImageFunction->EvaluateAtIndex(it.GetIndex())); |
| 490 |
|
} |
| 491 |
|
++it; |
| 492 |
|
++itN; |
| 493 |
|
} |
| 494 |
|
it.NextLine(); |
| 495 |
|
itN.NextLine(); |
| 496 |
|
} |
| 497 |
|
|
| 498 |
|
// Do the convolution in other directions |
| 499 |
LEN |
for(unsigned int direction=1;direction<itkGetStaticConstMacro(ImageDimension);direction++) |
| 500 |
|
{ |
| 501 |
|
|
| 502 |
|
size[direction] = 1; |
| 503 |
|
ind = centerIndex; |
| 504 |
|
for(unsigned int i = 0;i<itkGetStaticConstMacro(ImageDimension);i++) |
| 505 |
|
{ |
| 506 |
|
if(i > direction) |
| 507 |
|
{ |
| 508 |
|
ind[i] = 0; |
| 509 |
|
} |
| 510 |
|
} |
| 511 |
|
region.SetSize(size); |
| 512 |
|
region.SetIndex(ind); |
| 513 |
|
|
| 514 |
|
|
| 515 |
|
m_OperatorImageFunction->SetInputImage(m_InternalImage); |
| 516 |
|
m_OperatorImageFunction->SetOperator(m_ContinuousOperatorArray[direction]); |
| 517 |
|
|
| 518 |
LEN |
itk::ImageLinearIteratorWithIndex<InternalImageType> it(m_InternalImage,region); |
| 519 |
|
|
| 520 |
|
unsigned int dir = direction +1; |
| 521 |
|
if(dir == itkGetStaticConstMacro(ImageDimension)) |
| 522 |
|
{ |
| 523 |
|
dir = itkGetStaticConstMacro(ImageDimension)-1; |
| 524 |
|
} |
| 525 |
|
|
| 526 |
|
it.SetDirection(dir); |
| 527 |
|
it.GoToBeginOfLine(); |
| 528 |
|
while(!it.IsAtEnd()) |
| 529 |
|
{ |
| 530 |
|
while(!it.IsAtEndOfLine()) |
| 531 |
|
{ |
| 532 |
|
it.Set(m_OperatorImageFunction->EvaluateAtIndex(it.GetIndex())); |
| 533 |
|
++it; |
| 534 |
|
} |
| 535 |
|
it.NextLine(); |
| 536 |
|
} |
| 537 |
|
} |
| 538 |
|
|
| 539 |
|
return m_InternalImage->GetPixel(centerIndex); |
| 540 |
|
} |
| 541 |
|
|
| 542 |
|
/** Evaluate the function at specified ContinousIndex position.*/ |
| 543 |
|
template <class TInputImage,class TOutput> |
| 544 |
|
TOutput |
| 545 |
|
GaussianBlurImageFunction<TInputImage,TOutput> |
| 546 |
|
::EvaluateAtContinuousIndex( const ContinuousIndexType & index ) const |
| 547 |
|
{ |
| 548 |
|
PointType point; |
| 549 |
|
for(unsigned int i=0; i<itkGetStaticConstMacro(ImageDimension);i++) |
| 550 |
|
{ |
| 551 |
|
point[i] = index[i]; |
| 552 |
|
} |
| 553 |
|
return this->Evaluate(point); |
| 554 |
|
|
| 555 |
|
} |
| 556 |
|
|
| 557 |
|
} // namespace itk |
| 558 |
|
|
| 559 |
|
#endif |
| 560 |
|
|