[vtkusers] running out of memory for vtkImageData

Prashanth prashanth.dumpuri at gmail.com
Wed May 4 11:44:05 EDT 2011


I was not aware of the  /LARGEADDRESSAWARE link flag and the /3GB boot
switch in XP. I will try those. P.S. I wish I were a computer
scientist/programmer. Unfortunately I'm just an engineer.
Thanks


On Wed, May 4, 2011 at 10:27 AM, John Drescher <drescherjm at gmail.com> wrote:

> On Wed, May 4, 2011 at 11:15 AM, Prashanth <prashanth.dumpuri at gmail.com>
> wrote:
> > All,
> >   I'm trying to allocate a vtkImageData of size 512 x 512 x 218 with 3
> > scalar components. To save memory I'm storing the scalar values as
> 'short',
> > but 'am still running out of memory. I'm running my code on a Windows 32
> bit
> > machine. All the previous posts about memory allocation for vtkImageData
> > have been answered with "switch to Windows 7 or a 64-bit machine".
> > Unfortunately neither of these solutions work for me. Any other
> suggestions?
> > Thanks
>
> It will be hard to use such a large image in a 2GB address space that
> is fragmented such that ~1.2GB is the largest memory allocation you
> can make regardless of how much ram or swap you have.
>
> Did you add the /LARGEADDRESSAWARE link flag to your settings and use
> the /3GB boot switch in XP or use Vista and above?
>
> This will extend the limit to 3GB and reduce the fragmentation
> problem. Even with this I had serious problems processing my 0.625 mm
> LungCT data. If you are stuck with this avoid copying the image buffer
> and try to avoid any image type other than 1 component short.
> Originally I used a vtkBlender to combine a color overlay with my
> LungCT and that ended up using almost 800 MB per view (Axial, Coronal,
> Sagittal) I replaced that with separate actors for the overlay and
> image and that cut the memory usage down to less than 1/3 so now I can
> process 1 case without running into problems. Although when I now
> think of it I remember a developer recently telling me if I restricted
> the extent of the blending operation to a slice it would have reduced
> the memory footprint down to just the memory needed for blending.
>
> John
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20110504/eee9830f/attachment.htm>


More information about the vtkusers mailing list