How to use Image filter on series images

Beginner Questions
1

Hallo everyone

All example i can found about Morphological Image filter are base on single image. I tried some of them it can work on single image of my dataset.

But, Now i import a dicom series and i want to do the same processing on every slice,But it didn`t work. Or can say it run in bad way ,the outputimage is strange.
Here is my image after segmentation,you can see the edge of object is rough. so i want to simplely use opening opreation to smooth the edge. But when I used it, the image got weird.

o1
o1514×509 1.48 KB

w1
Code part:

    const unsigned int Dimension = 3;
using PixelType = unsigned short;
using ImageType = itk::Image< PixelType, Dimension >;
using RealPixelType = unsigned char;
using BinarImageType = itk::Image< RealPixelType, Dimension>;
using ReaderType = itk::ImageSeriesReader< ImageType >;
using ImageIOType = itk::GDCMImageIO;
using InputNamesGeneratorType = itk::GDCMSeriesFileNames;

ImageIOType::Pointer gdcmImageIO = ImageIOType::New();
ReaderType::Pointer reader = ReaderType::New();
InputNamesGeneratorType::Pointer inputNames = InputNamesGeneratorType::New();
inputNames->SetUseSeriesDetails(true);
inputNames->SetGlobalWarningDisplay(false);
inputNames->SetInputDirectory(InputPath);
    InputPath=""
const ReaderType::FileNamesContainer & filenames =
inputNames->GetInputFileNames();
reader->SetImageIO(gdcmImageIO);
reader->SetFileNames(filenames);

std::cout << "The input series in directory " << InputPath
	<< " has " << filenames.size() << " files" << std::endl;

try
{
	reader->Update();
}
catch (itk::ExceptionObject &excp)
{
	std::cerr << "Exception thrown while reading the series" << std::endl;
	std::cerr << excp << std::endl;
	return EXIT_FAILURE;
}
if (InputPath !="") {
	std::cout << "Path：" << InputPath;
}

//////
using FilterType = itk::OtsuThresholdImageFilter<ImageType, ImageType >;
FilterType::Pointer filter = FilterType::New();
filter->SetInput(reader->GetOutput());
try
{
	filter->Update();
}
catch (itk::ExceptionObject & excp)
{
	std::cerr << "Exception thrown " << excp << std::endl;
}
int threshold = filter->GetThreshold();
std::cout << "Threshold = " << threshold << std::endl;
//////////////////
using StructuringElementType = itk::BinaryBallStructuringElement<RealPixelType, 3>;
StructuringElementType structuringElement;
structuringElement.SetRadius(1);
structuringElement.CreateStructuringElement();

typedef itk::CastImageFilter< ImageType, BinarImageType >    CastToRealFilterType;
typedef itk::RescaleIntensityImageFilter< BinarImageType, ImageType > RescaleFilterType;
CastToRealFilterType::Pointer toReal = CastToRealFilterType::New();
RescaleFilterType::Pointer    rescale = RescaleFilterType::New();

toReal->SetInput(filter->GetOutput());
toReal->Update();

using BinaryMorphologicalOpeningImageFilterType =
	itk::BinaryMorphologicalOpeningImageFilter<BinarImageType, BinarImageType, StructuringElementType>;
BinaryMorphologicalOpeningImageFilterType::Pointer openingFilter = BinaryMorphologicalOpeningImageFilterType::New();
openingFilter->SetInput(toReal->GetOutput());
openingFilter->SetKernel(structuringElement);
openingFilter->Update();
rescale->SetInput(openingFilter->GetOutput());
rescale->Update();

  ...display in vtk

I think the reason is my StructuringElementType has 3 Dimension and then OpeningImageFilter will compute it in wrong way. but it not be allowed to change it to 2 Because my dataset has 3 Dimension.

i want to konw how to apply it in right way.Or how about treating each image layer separately?
This is very important to me. Thank you in advance.

2

If you want to apply an operation to each slice independently (and not in a 3D fashion), you could use SliceBySliceImageFilter.

Alternatively, have a for loop of 2D processing. Current pseudo-code:

read all slices
process 3D
visualize

could be changed to:

for all slices:
  read slice
  process 2D
  visualize
1 Like
3

That is what i want!! Thank you again!