[Insight-developers] itkMultiResolutionPDEDeformableRegistrationTest failures

[Bradley Lowekamp]

I took a quick look at this test. What I am finding interesting is  
that internally the filter compares the fixed and warped moving image.  
On most of the test that are failing it produces:

Compare warped moving and fixed.
Number of pixels different: 2
While on many  that are passing ( some do report only 2 ) it produces:
Compare warped moving and fixed.
Number of pixels different: 776
This output does not seem to make since with the comparison to the  
baseline image. Oddness... The error image seems to be off more then  
just one pixel due to rounding during resampling. I am unfortunately  
of no help with what may be causing these things.
Brad

On Aug 20, 2009, at 12:33 PM, Bill Lorensen wrote:

> Luis,
>
> The test itkMultiResolutionPDEDeformableRegistrationTest is failing on
> many platforms. Each failing platform has the same image error. We
> should check in a second regression image for
> itkMultiResolutionPDEDeformableRegistrationTestPixelCentered.png.
>
> As Tom V. pointed out recently, the nearest neighbor calculations can
> vary on some machines due to floating point precision.
>
> Can you check in a .1.png image from one of the failing platforms. All
> of my systems pass.
>
> Bill
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers

========================================================
Bradley Lowekamp
Lockheed Martin Contractor for
Office of High Performance Computing and Communications
National Library of Medicine
blowekamp at mail.nih.gov


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20090820/22a2d77b/attachment.htm>