[vtkusers] Problem with vtkPolyDataToImageStencil
Dženan Zukić
dzenanz at gmail.com
Fri Jul 27 08:44:48 EDT 2012
I have switched from ITK to VTK voxelizer, due to a bug in
ITK<https://issues.itk.org/jira/browse/ITK-2882>
.
On Fri, Jul 27, 2012 at 9:52 AM, Roman Grothausmann <
roman.grothausmann at helmholtz-berlin.de> wrote:
> Hello Matheus,
>
>
> I've experienced the same problem, and thanks to David I've now a more
> clear understanding when this happens. So far I've only checked if the mesh
> is closed. I will add a check with tolerance >0.
> With colleagues I've prepared a blender plug-in based on vtk filters for
> voxelizing blender meshes. It will be published on Midas as soon as my
> colleagues have given their OK. Let me know if You'd be interesting in
> testing the script before.
> Otherwise You could try vtkPartialVolumeModeller from Cory Quammen and
> Taylor II R.M.: http://www.insight-journal.**org/browse/publication/792<http://www.insight-journal.org/browse/publication/792>
> Corry is just upgrading the code a bit.
>
> One other option I know of is to use ITK, which during my testing, was not
> quicker than the VTK approach with vtkPolyDataToImageStencil. Below is a
> little test program.
>
> David: A more robust vtkPolyDataToImageStencil would be really great! Let
> me know if there is any upgrade to test.
>
> Happy coding
> Roman
>
> ////////program to voxelize a vtk-polydata-mesh into a itk-voxel-image
> ////////supposingly quicker than pure vtk: http://www.cmake.org/Wiki/VTK/*
> *Examples/PolyDataToImageData<http://www.cmake.org/Wiki/VTK/Examples/PolyDataToImageData>
> //allocating image before voxelization
>
> //#include <itkImage.h>
> //#include <itkQuadEdgeMesh.h>
> #include <itkVTKPolyDataReader.h> //can only read ASCII legacy VTK-files
> containing only triangles!!!
> #include <**itkTriangleMeshToBinaryImageFi**lter.h>
> #include <itkImageFileWriter.h>
> #include "itkFilterWatcher2.h"
>
>
> int main( int argc, char *argv[] ){
>
>
> if( argc != 9 )
> {
> std::cerr << "Usage: " << argv[0];
> std::cerr << " input_VTKpolydata-mesh";
> std::cerr << " outputImage";
> std::cerr << " dx dy dz ox oy oz";
> std::cerr << std::endl;
> return EXIT_FAILURE;
> }
>
>
>
> typedef unsigned char PixelType;
> const unsigned int Dimension = 3;
>
> //typedef itk::DefaultDynamicMeshTraits<**double, 3, 3,double,double>
> TriangleMeshTraits;
> //typedef itk::Mesh<double,3, TriangleMeshTraits> TriangleMeshType;
>
> //typedef itk::QuadEdgeMesh< float, Dimension > MeshType;
>
> typedef itk::Mesh< double, 3 > MeshType;
>
> typedef itk::Image< PixelType, Dimension > ImageType;
>
>
> typedef itk::VTKPolyDataReader<**MeshType> VTKmeshreaderType;
>
>
> VTKmeshreaderType::Pointer meshReader = VTKmeshreaderType::New();
> meshReader->SetFileName(argv[**1]);
>
> std::cout << "Reading: " << argv[1] << std::endl;
> try
> {
> meshReader->Update();
> }
> catch( itk::ExceptionObject & excp )
> {
> std::cerr << "Exception: " << excp << std::endl;
> return EXIT_FAILURE;
> }
>
> std::cout << "Read: " << argv[1] << std::endl;
>
>
> // Set Size, Spacing and origin
> ImageType::SizeType size;
> size[ 0 ] = atoi(argv[3]);
> size[ 1 ] = atoi(argv[4]);
> size[ 2 ] = atoi(argv[5]);
>
> ImageType::SpacingType spacing;
> spacing[0] = 1; //100.0 / size[0];
> spacing[1] = 1; //100.0 / size[1];
> spacing[2] = 1; //100.0 / size[2];
>
> ImageType::PointType origin;
> origin[0]= atoi(argv[6]);
> origin[1]= atoi(argv[7]);
> origin[2]= atoi(argv[8]);
>
> //allocate the output image
> ImageType::Pointer output = ImageType::New();
>
> output->SetRegions(size);
> output->SetSpacing(spacing);
> output->SetOrigin(origin);
> //output->SetRegions(**meshReader->GetOutput()->**
> GetRequestedRegion());
> output->Allocate();
>
> std::cout << "Image allocated!" << std::endl;
>
> //Set Inside/Outside voxel value
> const PixelType empty = 0;
> const PixelType fill = 255;
>
>
>
> typedef itk::**TriangleMeshToBinaryImageFilte**r<MeshType, ImageType>
> MeshFilterType;
> MeshFilterType::Pointer meshFilter = MeshFilterType::New();
>
> meshFilter->SetInput(**meshReader->GetOutput());
> meshFilter->SetInfoImage(**output); //
> meshFilter->SetTolerance (1.0);
> //meshFilter->SetSize (size);
> //meshFilter->SetSpacing (spacing);
> //meshFilter->SetOrigin(**origin);
> //meshFilter->SetIndex (index);
> //meshFilter->**SetUseObjectValue( true );
> meshFilter->SetInsideValue(**fill);
> meshFilter->SetOutsideValue(**empty);
> FilterWatcher watcher(meshFilter, "filter");
> meshFilter->Update();
>
> // Write the image
> typedef itk::ImageFileWriter< ImageType > WriterType;
> WriterType::Pointer writer = WriterType::New();
>
> writer->SetFileName(argv[2]);
> writer->SetInput(meshFilter->**GetOutput());
> try
> {
> meshFilter->Update();
> writer->Update();
> }
> catch( itk::ExceptionObject & excp )
> {
> std::cerr << excp << std::endl;
> return EXIT_FAILURE;
> }
>
>
> return EXIT_SUCCESS;
>
> }
>
>
>
> On 27.07.2012 00:53, David Gobbi wrote:
>
>> Hi Matheus,
>>
>> The vtkPolyDataToImageStencil filter is known to produce streaks like
>> those under these conditions:
>> 1) if the input data has free edges or is non-manifold
>> 2) if the distance between points is small enough that it approaches
>> the numerical tolerance of 32-bit floats
>>
>> You can check the first condition with vtkFeatureEdges, using the
>> following code:
>>
>> vtkFeatureEdges *edges = vtkFeatureEdges::New();
>> edges->SetInputConnection(**yourdata->GetOutputPort());
>> edges->FeatureEdgesOff();
>> edges->NonManifoldEdgesOn();
>> edges->BoundaryEdgesOn();
>> edges->Update();
>> cout << edges->GetOutput()->**GetNumberOfCells() << endl;
>>
>> It should print "0" if your data is nice, well-defined closed surface.
>>
>> You can check the second condition with vtkCleanPolyData. If your
>> polydata is changed by vtkCleanPolyData when the tolerance is set to
>> 1e-7, then vtkPolyDataToImageStencil might give incorrect results due
>> to roundoff error.
>>
>> I have plans to improve the vtkPolyDataToImageStencil code to make
>> it more robust, but it will be several months (at least) before it rises
>> to the top of my to-do list.
>>
>> - David
>>
>>
>> On Thu, Jul 26, 2012 at 3:32 PM, matheus_viana <vianamp at gmail.com> wrote:
>>
>>> Hello guys.
>>>
>>> I've a polydata and I'd like to voxelize that in order to obtain an
>>> ImageData. In order to do so, I'm using the following code:
>>>
>>> vtkImageData *blankImage = vtkImageData::New();
>>> blankImage -> SetExtent(extent);
>>> blankImage -> SetOrigin(0,0,0);
>>> blankImage -> SetSpacing(1,1,1);
>>> blankImage -> SetScalarTypeToUnsignedChar();
>>> blankImage -> AllocateScalars();
>>>
>>> vtkPolyDataToImageStencil *pol2Stenc = vtkPolyDataToImageStencil::**
>>> New();
>>> pol2Stenc -> SetTolerance(0.5);
>>> pol2Stenc -> SetInput(myPolyData);
>>> pol2Stenc -> SetInformationInput(**blankImage);
>>> pol2Stenc -> Update();
>>>
>>> vtkImageStencil *stencil = vtkImageStencil::New();
>>> stencil -> SetInput(blankImage);
>>> stencil -> ReverseStencilOff();
>>> stencil -> SetStencil(pol2Stenc->**GetOutput());
>>> stencil -> Update();
>>> vtkImageData *image = stencil -> GetOutput();
>>>
>>> It does the job, but I've observed some artifacts in the final image, as
>>> can
>>> be seen in the attached picture.
>>>
>>> Dos anyone have any idea about how to solve this problem? Is that a bug?
>>> Or
>>> am I doing any stupid thing?
>>>
>>> Cheers
>>> Matheus
>>>
>>> http://vtk.1045678.n5.nabble.**com/file/n5714947/Viana-**
>>> vtkPolyDataToImageStencil.jpg<http://vtk.1045678.n5.nabble.com/file/n5714947/Viana-vtkPolyDataToImageStencil.jpg>
>>>
>>>
>>>
>>> --
>>> View this message in context: http://vtk.1045678.n5.nabble.**
>>> com/Problem-with-**vtkPolyDataToImageStencil-**tp5714947.html<http://vtk.1045678.n5.nabble.com/Problem-with-vtkPolyDataToImageStencil-tp5714947.html>
>>> Sent from the VTK - Users mailing list archive at Nabble.com.
>>> ______________________________**_________________
>>> Powered by www.kitware.com
>>>
>>> Visit other Kitware open-source projects at http://www.kitware.com/**
>>> opensource/opensource.html<http://www.kitware.com/opensource/opensource.html>
>>>
>>> Please keep messages on-topic and check the VTK FAQ at:
>>> http://www.vtk.org/Wiki/VTK_**FAQ <http://www.vtk.org/Wiki/VTK_FAQ>
>>>
>>> Follow this link to subscribe/unsubscribe:
>>> http://www.vtk.org/mailman/**listinfo/vtkusers<http://www.vtk.org/mailman/listinfo/vtkusers>
>>>
>> ______________________________**_________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at http://www.kitware.com/**
>> opensource/opensource.html<http://www.kitware.com/opensource/opensource.html>
>>
>> Please keep messages on-topic and check the VTK FAQ at:
>> http://www.vtk.org/Wiki/VTK_**FAQ <http://www.vtk.org/Wiki/VTK_FAQ>
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/**listinfo/vtkusers<http://www.vtk.org/mailman/listinfo/vtkusers>
>>
>>
> --
> Roman Grothausmann
>
> Helmholtz-Zentrum Berlin für Materialien und Energie GmbH
> Bereich Funktionale Materialien
> Institut für angewandte Materialforschung
> Hahn-Meitner-Platz 1
> D-14109 Berlin
>
> (früher Hahn-Meitner-Institut und BESSY)
>
>
> Tel.: +49-(0)30-8062-42816
> Fax.: +49-(0)30-8062-43059
>
> Vorsitzender des Aufsichtsrats: Prof. Dr. Dr. h.c. mult. Joachim Treusch
> Stellvertretende Vorsitzende: Dr. Beatrix Vierkorn-Rudolph
> Geschäftsführer: Prof. Dr. Anke Rita Kaysser-Pyzalla, Dr. Ulrich Breuer
> Sitz der Gesellschaft: Berlin
> Handelsregister: AG Charlottenburg, 89 HRB 5583
>
>
>
>
> ______________________________**_________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/**
> opensource/opensource.html<http://www.kitware.com/opensource/opensource.html>
>
> Please keep messages on-topic and check the VTK FAQ at:
> http://www.vtk.org/Wiki/VTK_**FAQ <http://www.vtk.org/Wiki/VTK_FAQ>
>
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/**listinfo/vtkusers<http://www.vtk.org/mailman/listinfo/vtkusers>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20120727/24b6b3a1/attachment.htm>
More information about the vtkusers
mailing list