[ITK-users] Buffered region of size [0, 0, 0]

[Bastian A.]

To answer my own question:

the issue was that one of the filters upstream had set InPlaceOn(), which
caused one of the inputs require later to be destroyed.
See
http://public.kitware.com/pipermail/insight-users/2011-November/043084.html
for details.

Bastian

On Mon, Apr 20, 2015 at 3:45 PM, Bastian A. <bastianra at gmail.com> wrote:

> Hi,
>
> I am trying to multiply two images in a loop. Calling Print() on the
> inputs of the MultiplyImageFilter seems to indicate that they have the
> same LargestPossible, Buffered and Requested regions.
> However, calling UpdateLargestPossibleRegion on the MultiplyImageFilter
> will result in the failed assertion during the first iteration:
>
> Assertion `Region ImageRegion (0x7fff777f5b80)
>   Dimension: 3
>   Index: [-95, -171, -52]
>   Size: [188, 675, 10]
>  is outside of buffered region ImageRegion (0xdec558)
>   Dimension: 3
>   Index: [0, 0, 0]
>   Size: [0, 0, 0]
> ' failed.
>
> I am completely lost here, why would the buffered region be of size [0, 0,
> 0]?
> My assuption was that AllocateOutputs() in the filter I am calling would
> take care of setting the requested region; both inputs seem to have the
> correct regions set.
>
> Any pointers on how to debug this issue would be great.
>
> Thanks,
> Bastian
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/insight-users/attachments/20150420/e6e4441e/attachment.html>