[ITK] itk 4.5.1 crashing in HDF5

[Matt McCormick]

Hi Chris,

How is DCMTK related?  Is ITK built with DCMTK enabled?  Is DCMTK
installed on the system? Is a system DCMTK available?

Where does the TIFF error message come from?

Thanks,
Matt

On Fri, May 9, 2014 at 2:56 PM, Chris Cooper
<chris.cooper at veranmedical.com> wrote:
> This is running on Windows 7 in a Release build (CMAKE_BUILD_TYPE  set to
> RelWithDebInfo)
> I’m not sure how to use a system version of HDF5, would I need to re-build
> itk with HDF5 somehow left out?
> It appears to be just this one file.  The MHD follows, sharing the RAW would
> be trickier …
>
> Here’s a strange thing I just noticed.  Sometimes the program runs fine, in
> those cases I am getting an error "Not a TIFF file, bad magic number 25167
> (0x624f).” - it’s when I don’t get that error that I get the crash instead.
>
> We are doing some testing that makes it seem like dcmtk is the problem, if
> we use the official 3.6.0 release instead of the snapshot from last
> November, the problem goes away …
>
> Thanks,
> Chris
>
> MHD file looks like this:
>
> ObjectType = Image
> NDims = 3
> BinaryData = True
> BinaryDataByteOrderMSB = False
> CompressedData = False
> TransformMatrix = 1 0 0 0 1 0 0 0 1
> Offset = 0 0 0
> CenterOfRotation = 0 0 0
> AnatomicalOrientation = RAI
> ElementSpacing = 0.615234 0.615234 0.5
> DimSize = 512 512 540
> ElementType = MET_SHORT
> ElementDataFile = media-IO.20140509.1233344180.n540.recon.itks163d.raw
>
>
> From: Bradley Lowekamp <blowekamp at mail.nih.gov>
> Date: Friday, May 9, 2014 at 12:30 PM
> To: Chris Cooper <chris.cooper at veranmedical.com>
> Cc: "community at itk.org" <community at itk.org>
> Subject: Re: [ITK] itk 4.5.1 crashing in HDF5
>
> Hello,
>
> What is you platform, architecture and are you using the ITK provided HDF5?
> How are you compiling ITK Debug Release?
>
> If you use a system version of HDF5 do you still get the issue?
>
> Is there an issue will all HDF5 files your just this one? Can you share the
> problematic file?
>
> Brad
>
>
> On May 9, 2014, at 2:23 PM, Chris Cooper <chris.cooper at veranmedical.com>
> wrote:
>
> While trying to read a file, we get a crash with this call stack:
> H5MM_xfree
> H5E_printf_stack
> H5E_clear_stack
> H5Fis_hdf5
> H5::H5File::isHdf5
> HDF5ImageIO::CanReadFile
> ImageIOFactory::CreateImageIO
> ImageFileReader<…>::GenerateOutputInformation
> ProcessObject::UpdateOutputInformation
> ImageBase<3>::UpdateOutputInformation
> DataObject::Update
> ProcessObject::Update
>
> The code that calls itk looks like this:
> ReaderType::Pointer reader = ReaderType::New();
>         reader->DebugOff();
>         reader->SetGlobalWarningDisplay(false);
>         reader->SetFileName(cache.pathForUid(uid).string());
>         reader->Update();
>
> This looks similar to an issue discussed here:
> http://www.itk.org/pipermail/insight-users/2013-November/049157.html
>
> Any thoughts on this problem?  Is there a chance it’s been solved in 4.5.2
> (which I am downloading & building now)?
>
> Thanks,
> Chris
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/community
>
>
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/cgi-bin/mailman/listinfo/community
>