[Insight-developers] Compile error with UsePhysicalSpace in BinaryFunctorImageFilter

Williams, Norman K norman-k-williams at uiowa.edu
Thu Jun 2 10:52:45 EDT 2011 

I don't see the need to use ResampleImageFilter.  How is using an
Interpolator in the filter different from calling ResampleImageFilter?
Doesn't it just use an Interpolator internally?


On 6/2/11 9:43 AM, "Gaëtan Lehmann" <gaetan.lehmann at jouy.inra.fr> wrote:

>
>Le 2 juin 11 à 15:02, Luis Ibanez a écrit :
>
>> Kent,
>>
>> I think that we should back up the changes to the
>> BinaryFunctorImageFunctor and instead create a
>> helper filter that can do the equivalent resampling
>> on demand.
>>
>> Looking back at the recent changes in the Binary
>> FunctorImageFilter, what we did, was to put an
>> simplified version of the ResampleImageFilter
>> inside of the BinaryFunctor ImageFilter, to deal
>> with something that the ResampleImageFitler
>> can do from the outside.
>>
>
>Kent's feature don't need to allocate a new image like
>ResampleImageFilter would do.
>And of course it doesn't require the usage of an additional filter.
>
>> We can back up those changes, and instead
>> create a filter that will use the Resample image
>> filter to produce the chunks needed by the
>> BinaryFunctor image filter *from the outside*.
>>
>> I'll be happy to provide such filter.
>>
>
>I agree that a check for the correct space position in
>BinaryFunctorImageFilter and that new filter would be a better way to
>fix that use case, even if it requires more memory and the usage of an
>additional filter.
>
>The check is a backward compatibility issue which will be seen at
>runtime only though.
>
>Gaëtan
>
>
>
>>
>>     Luis
>>
>>
>> --------------------------------------------------------------
>> 2011/6/2 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>:
>>>
>>> Kent,
>>>
>>> There is another problem: the interpolators are in another module
>>> which is
>>> not required by ITK-ImageFilterBase where those filters are.
>>> I'm not sure that's a good idea to add that dependency.
>>>
>>> Luis, Xiaoxiao, what do you think?
>>>
>>> The other problem is that I'm still not sure that
>>> BinaryFunctorImageFilter
>>> is still able to build with any pixel type as it was before,
>>> because the
>>> interpolators may have some assumptions on the pixel type that
>>> BinaryFunctorImageFilter doesn't have.
>>> For example I think it was possible to make
>>> BinaryFunctorImageFilter compile
>>> with itk::Image< std::list< int >, 3 > as image type - is it still
>>> possible?
>>> I think that feature is important enough to tolerate some loss in
>>> possible
>>> pixel types, but something must be done about backward compatibility.
>>>
>>> Gaëtan
>>>
>>>
>>>
>>> Le 1 juin 11 à 22:18, Williams, Norman K a écrit :
>>>
>>>> Yes, implementing GenerateInputRequestedRegion does the trick. New
>>>> patch
>>>> pushed.
>>>>
>>>>
>>>> The test passes now without exceptions.  The one irksome issue is
>>>> when
>>>> adding (or subtracting or whatever) images where the interpolated
>>>> image is
>>>> smaller. The point derived from Input1's index is reported outside
>>>> the
>>>> smaller Input2 image by the interpolater, on edge voxels.  I'm
>>>> sure this
>>>> has to do with the voxel centers not lining up.
>>>>
>>>> I'm not sure exactly what to do about that or whether I need to do
>>>> anything ;-)
>>>>
>>>> On 6/1/11 1:46 PM, "Gaëtan Lehmann" <gaetan.lehmann at jouy.inra.fr>
>>>> wrote:
>>>>
>>>>>
>>>>> Kent,
>>>>>
>>>>> Thanks for pushing the change to gerrit. I've updated it with a
>>>>> small
>>>>> fix for DivideImageFilter. It builds now.
>>>>>
>>>>> For the requested region, I think you should reimplement
>>>>> GenerateInputRequestedRegion() and request the largest possible
>>>>> region
>>>>> of the 2nd (and 3rd) inputs if UsePhysicalSpace is true and the
>>>>> origin, orientation and spacing are not matching the first input.
>>>>>
>>>>> Regards,
>>>>>
>>>>> Gaëtan
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Le 1 juin 11 à 19:36, Gaëtan Lehmann a écrit :
>>>>>
>>>>>>
>>>>>> Le 1 juin 11 à 18:52, Williams, Norman K a écrit :
>>>>>>
>>>>>>> I have a fix for this, but I have larger problems with these
>>>>>>> filters, as
>>>>>>> detailed in my last message to the ITK list.
>>>>>>>
>>>>>>>
>>>>>>> The problem being, once I correct bugs in the test, the filters
>>>>>>> start
>>>>>>> failing.  My changes to the filters, as written, will fail
>>>>>>> because
>>>>>>> of size
>>>>>>> mismatches in the input images. This, unfortunately is one of the
>>>>>>> things
>>>>>>> my changes were explicitly supposed to support!
>>>>>>
>>>>>> Please push your changes to gerrit so we can look at this last
>>>>>> problem.
>>>>>>
>>>>>> Gaëtan
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>> On 5/31/11 3:25 PM, "Gaëtan Lehmann"
>>>>>>> <gaetan.lehmann at jouy.inra.fr>
>>>>>>> wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> Hi Kent,
>>>>>>>>
>>>>>>>> I get a compilation error in WrapITK with your recent changes in
>>>>>>>> BinaryFunctorImageFilter and complex pixel types - see the
>>>>>>>> error at
>>>>>>>> the end of the message.
>>>>>>>> I'm not so sure, but I think the problem comes from the lack of
>>>>>>>> casting from the output type of the interpolator to the pixel
>>>>>>>> type
>>>>>>>> of
>>>>>>>> the second input. Or maybe the output of the interpolator is
>>>>>>>> not of
>>>>>>>> the right type?
>>>>>>>> BTW, is it possible to make the interpolator work with any
>>>>>>>> kind of
>>>>>>>> pixel? The interpolator may do some assumptions on the pixel
>>>>>>>> types
>>>>>>>> which are not done in BinaryFunctorImageFilter...
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>>
>>>>>>>> Gaëtan
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> In file included from /home/glehmann/src/ITK/Modules/Filtering/
>>>>>>>> ImageFilterBase/include/itkBinaryFunctorImageFilter.h:267,
>>>>>>>>              from /home/glehmann/src/ITK/Modules/Filtering/
>>>>>>>> ImageCompare/include/itkAbsoluteValueDifferenceImageFilter.h:22,
>>>>>>>>              from /home/glehmann/src/ITK/build/Wrapping/WrapITK/
>>>>>>>> Typedefs/PixelMathExplicit.h:38,
>>>>>>>>              from /home/glehmann/src/ITK/build/Wrapping/WrapITK/
>>>>>>>> Libraries/PixelMath/itkMultiplyImageFilterExplicit.cxx:3:
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageFilterBase/
>>>>>>>> include/
>>>>>>>> itkBinaryFunctorImageFilter.txx: In member function 'void
>>>>>>>> itk::BinaryFunctorImageFilter<TInputImage1, TInputImage2,
>>>>>>>> TOutputImage, TFunction>::ThreadedGenerateData(const typename
>>>>>>>> TOutputImage::RegionType&, int) [with TInputImage1 =
>>>>>>>> itk::Image<std::complex<float>, 3u>, TInputImage2 =
>>>>>>>> itk::Image<std::complex<float>, 3u>, TOutputImage =
>>>>>>>> itk::Image<std::complex<float>, 3u>, TFunction =
>>>>>>>> itk::Functor::Mult<std::complex<float>, std::complex<float>,
>>>>>>>> std::complex<float> >]':
>>>>>>>> /home/glehmann/src/ITK/build/Wrapping/WrapITK/Libraries/
>>>>>>>> PixelMath/
>>>>>>>> itkMultiplyImageFilterExplicit.cxx:21:   instantiated from here
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageFilterBase/
>>>>>>>> include/
>>>>>>>> itkBinaryFunctorImageFilter.txx:302: error: no match for call to
>>>>>>>> '(itk::Functor::Mult<std::complex<float>, std::complex<float>,
>>>>>>>> std::complex<float> >) (std::complex<float>,
>>>>>>>> std::complex<double>&)'
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageIntensity/include/
>>>>>>>> itkMultiplyImageFilter.h:58: note: candidates are: TOutput
>>>>>>>> itk::Functor::Mult<TInput1, TInput2, TOutput>::operator()(const
>>>>>>>> TInput1&, const TInput2&) const [with TInput1 =
>>>>>>>> std::complex<float>,
>>>>>>>> TInput2 = std::complex<float>, TOutput = std::complex<float>]
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageFilterBase/
>>>>>>>> include/
>>>>>>>> itkBinaryFunctorImageFilter.txx: In member function 'void
>>>>>>>> itk::BinaryFunctorImageFilter<TInputImage1, TInputImage2,
>>>>>>>> TOutputImage, TFunction>::ThreadedGenerateData(const typename
>>>>>>>> TOutputImage::RegionType&, int) [with TInputImage1 =
>>>>>>>> itk::Image<std::complex<float>, 2u>, TInputImage2 =
>>>>>>>> itk::Image<std::complex<float>, 2u>, TOutputImage =
>>>>>>>> itk::Image<std::complex<float>, 2u>, TFunction =
>>>>>>>> itk::Functor::Mult<std::complex<float>, std::complex<float>,
>>>>>>>> std::complex<float> >]':
>>>>>>>> /home/glehmann/src/ITK/build/Wrapping/WrapITK/Libraries/
>>>>>>>> PixelMath/
>>>>>>>> itkMultiplyImageFilterExplicit.cxx:21:   instantiated from here
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageFilterBase/
>>>>>>>> include/
>>>>>>>> itkBinaryFunctorImageFilter.txx:302: error: no match for call to
>>>>>>>> '(itk::Functor::Mult<std::complex<float>, std::complex<float>,
>>>>>>>> std::complex<float> >) (std::complex<float>,
>>>>>>>> std::complex<double>&)'
>>>>>>>> /home/glehmann/src/ITK/Modules/Filtering/ImageIntensity/include/
>>>>>>>> itkMultiplyImageFilter.h:58: note: candidates are: TOutput
>>>>>>>> itk::Functor::Mult<TInput1, TInput2, TOutput>::operator()(const
>>>>>>>> TInput1&, const TInput2&) const [with TInput1 =
>>>>>>>> std::complex<float>,
>>>>>>>> TInput2 = std::complex<float>, TOutput = std::complex<float>]
>>>>>>>> make[2]: *** [Wrapping/WrapITK/Libraries/PixelMath/CMakeFiles/
>>>>>>>> PixelMathExplicit.dir/itkMultiplyImageFilterExplicit.cxx.o]
>>>>>>>> Error 1
>>>>>>>> make[1]: *** [Wrapping/WrapITK/Libraries/PixelMath/CMakeFiles/
>>>>>>>> PixelMathExplicit.dir/all] Error 2
>>>>>>>> make[1]: *** Waiting for unfinished jobs....
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> Gaëtan Lehmann
>>>>>>>> Biologie du Développement et de la Reproduction
>>>>>>>> INRA de Jouy-en-Josas (France)
>>>>>>>> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>>>>>>>> http://voxel.jouy.inra.fr  http://www.itk.org
>>>>>>>> http://www.mandriva.org  http://www.bepo.fr
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Powered by www.kitware.com
>>>>>>>>
>>>>>>>> Visit other Kitware open-source projects at
>>>>>>>> http://www.kitware.com/opensource/opensource.html
>>>>>>>>
>>>>>>>> Kitware offers ITK Training Courses, for more information visit:
>>>>>>>> http://kitware.com/products/protraining.html
>>>>>>>>
>>>>>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>>>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>>>>>
>>>>>>>> Follow this link to subscribe/unsubscribe:
>>>>>>>> http://www.itk.org/mailman/listinfo/insight-developers
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> ________________________________
>>>>>>> Notice: This UI Health Care e-mail (including attachments) is
>>>>>>> covered by the Electronic Communications Privacy Act, 18 U.S.C.
>>>>>>> 2510-2521, is confidential and may be legally privileged.  If you
>>>>>>> are not the intended recipient, you are hereby notified that any
>>>>>>> retention, dissemination, distribution, or copying of this
>>>>>>> communication is strictly prohibited.  Please reply to the sender
>>>>>>> that you have received the message in error, then delete it.
>>>>>>> Thank
>>>>>>> you.
>>>>>>> ________________________________
>>>>>>
>>>>>> --
>>>>>> Gaëtan Lehmann
>>>>>> Biologie du Développement et de la Reproduction
>>>>>> INRA de Jouy-en-Josas (France)
>>>>>> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>>>>>> http://voxel.jouy.inra.fr  http://www.itk.org
>>>>>> http://www.mandriva.org  http://www.bepo.fr
>>>>>>
>>>>>> _______________________________________________
>>>>>> Powered by www.kitware.com
>>>>>>
>>>>>> Visit other Kitware open-source projects at
>>>>>> http://www.kitware.com/opensource/opensource.html
>>>>>>
>>>>>> Kitware offers ITK Training Courses, for more information visit:
>>>>>> http://kitware.com/products/protraining.html
>>>>>>
>>>>>> Please keep messages on-topic and check the ITK FAQ at:
>>>>>> http://www.itk.org/Wiki/ITK_FAQ
>>>>>>
>>>>>> Follow this link to subscribe/unsubscribe:
>>>>>> http://www.itk.org/mailman/listinfo/insight-developers
>>>>>
>>>>> --
>>>>> Gaëtan Lehmann
>>>>> Biologie du Développement et de la Reproduction
>>>>> INRA de Jouy-en-Josas (France)
>>>>> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>>>>> http://voxel.jouy.inra.fr  http://www.itk.org
>>>>> http://www.mandriva.org  http://www.bepo.fr
>>>>>
>>>>
>>>>
>>>>
>>>> ________________________________
>>>> Notice: This UI Health Care e-mail (including attachments) is
>>>> covered by
>>>> the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is
>>>> confidential and may be legally privileged.  If you are not the
>>>> intended
>>>> recipient, you are hereby notified that any retention,
>>>> dissemination,
>>>> distribution, or copying of this communication is strictly
>>>> prohibited.
>>>>  Please reply to the sender that you have received the message in
>>>> error,
>>>> then delete it.  Thank you.
>>>> ________________________________
>>>
>>> --
>>> Gaëtan Lehmann
>>> Biologie du Développement et de la Reproduction
>>> INRA de Jouy-en-Josas (France)
>>> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>>> http://voxel.jouy.inra.fr  http://www.itk.org
>>> http://www.mandriva.org  http://www.bepo.fr
>>>
>>>
>
>--
>Gaëtan Lehmann
>Biologie du Développement et de la Reproduction
>INRA de Jouy-en-Josas (France)
>tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>http://voxel.jouy.inra.fr  http://www.itk.org
>http://www.mandriva.org  http://www.bepo.fr
>



________________________________
Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is confidential and may be legally privileged.  If you are not the intended recipient, you are hereby notified that any retention, dissemination, distribution, or copying of this communication is strictly prohibited.  Please reply to the sender that you have received the message in error, then delete it.  Thank you.
________________________________

More information about the Insight-developers mailing list