[Tubetk-developers] HDCS Enhancing doesn't work

kenichiro yoshimi rccm.kyoshimi at gmail.com
Mon Nov 25 18:33:46 EST 2013


Hi Stephen,

Probably In my case, HDCS enhancing might work well with the modification
to itktubeAnisotropicHybridDiffusionImageFilter.hxx as described below.
But, I don't know whether this change is valid or not.

l. 327
 EigenVectorMatrixType eigenVectorMatrix;
  ==>
 EigenVectorMatrixType eigenVectorMatrix = eigenVectorImageIterator.Get();

Thanks a lot,
yoshimi


2013/11/25 kenichiro yoshimi <rccm.kyoshimi at gmail.com>

> Hi Stephen,
>
> Today, I could confirm that HDCS enhancing works, by setting the
> particular parameters (EED Contrast (1.0), CED Contrast (1.0) and Hybrid
> Contrast(1.0)) to testing case. Now I'm trying in my cases. Then I meet the
> failure to sort eigenvalues in
> "itktubeAnisotropicHybridDiffusionImageFilter.hxx". The following codes may
> be wrong:
>
> l. 216:
> if(  vnl_math_abs( eigenValue[i] > largest ) )
>   ==> if(  vnl_math_abs( eigenValue[i] ) > largest )
>
> l.226:
> if( eigenValue[i] != smallest && eigenValue[i] != largest )
>   ==> if( vnl_math_abs( eigenValue[i] ) != smallest &&
>              vnl_math_abs( eigenValue[i] ) != largest )
>
> Could you please confirm this?
>
> Thanks,
> yoshimi
>
>
> 2013/11/21 kenichiro yoshimi <rccm.kyoshimi at gmail.com>
>
>> Hi Stephen,
>>
>> Thank you for your kind advice.
>>
>> Following your instructions, I build a "Experimental" project, but the
>> tests may not run. Then I obtain the build log as:
>>
>> 1>------ Build : Project: Experimental, Release x64 ------
>> 1>   Site: YOSHIMI
>> 1>   Build name: Win32-vs9
>> 1>Create new tag: 20131120-2346 - Experimental
>> 1>Configure project
>> 1>   Each . represents 1024 bytes of output
>> 1>    . Size of output: 0K
>> 1>Build project
>> 1>   Each symbol represents 1024 bytes of output.
>> 1>   '!' represents an error and '*' a warning.
>> 1>    ......... Size of output: 8K
>> 1>   0 Compiler errors
>> 1>   0 Compiler warnings
>> 1>Test project G:/work/kitware/build/TubeTK
>> 1>No tests were found!!!
>> 1>Performing coverage
>> 1> Cannot find any coverage files. Ignoring Coverage request.
>> 1>Submit files (using http)
>> 1>   Using HTTP submit method
>> 1>   Drop site:http://open.cdash.org/submit.php?project=TubeTK
>> 1>   Uploaded:
>> G:/work/kitware/build/TubeTK/Testing/20131120-2346/Build.xml
>> 1>   Uploaded:
>> G:/work/kitware/build/TubeTK/Testing/20131120-2346/Configure.xml
>> 1>   Uploaded: G:/work/kitware/build/TubeTK/Testing/20131120-2346/Test.xml
>> 1>   Submission successful
>>
>> I am attaching the CMakeCache.txt file for the building TubeTK. I
>> appreciate if you would point out what I was doing wrong.
>>
>> Thanks,
>> yoshimi
>>
>>
>> 2013/11/20 Stephen Aylward <stephen.aylward at kitware.com>
>>
>>> Hi,
>>>
>>> Sorry that you are having troubles.
>>>
>>> When you build using windows, there should be a build target/project
>>> called "Experimental" that you can build from within Visual Studio
>>> 2008. Please try building that project.  It should cause all of the
>>> tests to run.  Please let me know if there are failures.
>>>
>>> Thanks,
>>> Stephen
>>>
>>>
>>> On Wed, Nov 20, 2013 at 6:36 AM, kenichiro yoshimi
>>> <rccm.kyoshimi at gmail.com> wrote:
>>> > Hi Stephen,
>>> >
>>> > Thanks for your reply.
>>> > I build TubeTK with Visual Studio 2008, but I don't know the way to
>>> run the
>>> > tests. So, I write the program that is  the same  as
>>> > "itktubeAnisotropicHybridDiffusionImageFilterTest.cxx" and run with the
>>> > testing data. The result has no error, but the output image data is
>>> the same
>>> > as input data just like using Slicer plugin.
>>> >
>>> > My Slicer versioin is 4.3.1-2013-11-07 r22692 64 bit. My system is
>>> Windows 8
>>> > 64bit.
>>> >
>>> > Best regards,
>>> > yoshimi
>>> >
>>> >
>>> > 2013/11/19 Stephen Aylward <stephen.aylward at kitware.com>
>>> >>
>>> >> Hi Kenichiro,
>>> >>
>>> >> Thank you for trying TubeTK.
>>> >>
>>> >> How did you install the TubeTK extensions?  If you compiled them, then
>>> >> I would like you to run the tests on your local build, and see if they
>>> >> pass.  As you said, we use that data in testing, and so if the tests
>>> >> pass, then the integration into Slicer is the problem.  If the tests
>>> >> fail, then the build of TubeTK is the problem.
>>> >>
>>> >> Also, what version of Slicer are you using?
>>> >>
>>> >> Thanks,
>>> >> Stephen
>>> >>
>>> >>
>>> >> On Mon, Nov 11, 2013 at 6:31 AM, kenichiro yoshimi
>>> >> <rccm.kyoshimi at gmail.com> wrote:
>>> >> > Dear all,
>>> >> >
>>> >> > I'm trying to use the Slicer plugin filter named "Hybrid Enhancing
>>> >> > Anisotropic Diffusion with Continuous Switch (TubeTK)". But it does
>>> not
>>> >> > reflect even if I specify various parameters. I try to apply it to
>>> >> > CroppedWholeLungCTScan.mhd which is used for testing framework.
>>> >> >
>>> >> > Can someone help me in getting this to work?
>>> >> >
>>> >> > Cheers,
>>> >> > yoshimi
>>> >> >
>>> >> > _______________________________________________
>>> >> > Tubetk-developers mailing list
>>> >> > Tubetk-developers at tubetk.org
>>> >> >
>>> http://public.kitware.com/cgi-bin/mailman/listinfo/tubetk-developers
>>> >> >
>>> >>
>>> >>
>>> >>
>>> >> --
>>> >>
>>> >> ====================================================
>>> >> Stephen R. Aylward, Ph.D.
>>> >> Senior Director of Operations, North Carolina, Kitware, Inc.
>>> >> http://www.kitware.com
>>> >> http://www.aylward.org
>>> >> (919) 969-6990 x300
>>> >> ====================================================
>>> >
>>> >
>>>
>>>
>>>
>>> --
>>>
>>> ====================================================
>>> Stephen R. Aylward, Ph.D.
>>> Senior Director of Operations, North Carolina, Kitware, Inc.
>>> http://www.kitware.com
>>> http://www.aylward.org
>>> (919) 969-6990 x300
>>> ====================================================
>>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/tubetk-developers/attachments/20131126/7729b1f1/attachment.html>


More information about the Tubetk-developers mailing list