[ITK] SCIFIO

[Mark Hiner]

Hi Michael,

When you get InvalidRequestedRegionError on a series after the first, it
strongly suggests series skew. In the dead-A.ome.tiff, series 2 is larger
(more planes) than series 1, and series 3 is smaller than series 2. So it
seems that at some point in the workflow the requested region is not
matching the current series... which doesn't cause any errors until the
region size decreases from one series to the next.

Note that without the intermediate Update() call on the Image, the program
will iterate successfully through the entire dataset. So it's the
intermediate Update() that's exposing this problem. I assume you do not
actually need it, since calling Update() on the writer will trigger the
whole pipeline and you can then extract the intermediate image.

On the other hand, I also assume you should be able to do incremental
Update() calls, but I have no idea why it's not working with the
SCIFIOImageIO. Perhaps someone else in the ITK community sees an error in
usage of the ITK API?

Otherwise I would suggest inspecting the requested region
<http://www.itk.org/Doxygen/html/classitk_1_1ImageBase.html#a14769ccb89e9830f8128e09bf022b174>
of various components at each series. A quick check of the Writer's region
shows it changing at each series appropriately. However, the region of your
image with intermediate updates never changes... but I don't really
understand why (since it's also downstream of the reader).

Sorry I can't be of more help,
Mark

On Tue, Oct 28, 2014 at 2:49 PM, Michael Meuli <michael.meuli at gmail.com>
wrote:

> Hi all
>
> If I modify the RunTest() function in itkSCIFIOImageIOTest.cxx just a
> little bit to
> get hold of the image for processing it, I get the following error
> which appears, depending on the file, after a
> certain number of series were processed with the following error:
>
> terminate called after throwing an instance of
> 'itk::InvalidRequestedRegionError'
>   what():
> /home/michael/bioimage/itk/itk-master/InsightToolkit/Modules/Core/Common/src/itkDataObject.cxx:393:
> Requested region is (at least partially) outside the largest possible
> region.
> Aborted (core dumped)
>
> In case of dead-A.ome.tiff after processing the first two series.
>
> Please find attached the runnable code.
> The file dead-A.ome.tiff can be found here:
> https://www.dropbox.com/sh/t2z0f4jttcnfh8s/AADq21HNR7EwH1JFxskhXXhea
> in the folder ome-tiff-files/5.0.5
>
> Googling for the poblem I found the following about the Update() fuction:
>
> http://www.itk.org/Doxygen/html/classitk_1_1ProcessObject.html#a4041fb21e9105500eee311e265691bd5
> I don't really understand it but I've tried
> UpdateLargestPossibleRegion() instead of Update() but the complier
> didn't eat it.
>
> Many thanks and best regards
>
> Michael
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/mailman/listinfo/community
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/community/attachments/20141029/803a6201/attachment.html>