[Insight-developers] Re: backward imcompatible changes in gaussian spatial object ?

[Julien Jomier]

Gaetan,

There was a bug in the itkGaussianSpatialObject. Basically it was not 
using the radius information and therefore was taking a very long time 
to be rasterized. If you want your test to pass with the current 
baseline image, you can increase the radius of the Gaussian and that 
should work, otherwise let me know.

Julien

Gaëtan Lehmann wrote:
> 
> Julien,
> 
> I seem that the changes you have done in the gaussian spatial objet 
> recently are changing the result produced by one the WrapITK tests:
> 
> http://www.itk.org/Testing/Sites/marvin.jouy.inra.fr/Linux-gcc-Debug/20070113-0100-Nightly/Results/__Wrapping_WrapITK_Python_Tests_PythonSpatialObject.html 
> 
> 
> Are those changes backward incompatibles, or is it a problem in the way 
> I have used that gaussian spatial object ?
> 
> Thanks,
> 
> Gaetan
> 
> 
> 
> import itk
> from sys import argv, stderr, exit
> 
> itk.auto_progress(2)
> 
> dim = 2
> SOType = itk.SpatialObject[dim]
> InternalImageType = itk.Image[itk.F, dim]
> OutputPixelType = itk.UC
> OutputImageType = itk.Image[OutputPixelType, dim]
> 
> ellipse = itk.EllipseSpatialObject[dim].New( Radius=[10,5] )
> ellipse.GetObjectToParentTransform().SetOffset( [20,20] )
> ellipse.ComputeObjectToWorldTransform()
> 
> box = itk.BoxSpatialObject[dim].New( Size=20 )
> box.GetObjectToParentTransform().SetOffset( [20,40] )
> box.ComputeObjectToWorldTransform()
> 
> gaussian = itk.GaussianSpatialObject[dim].New()
> gaussian.GetObjectToParentTransform().SetOffset( [60,60] )
> gaussian.GetObjectToParentTransform().SetScale( 10 )
> gaussian.ComputeObjectToWorldTransform()
> 
> group = itk.GroupSpatialObject[dim].New()
> group.AddSpatialObject( ellipse.GetPointer() )
> group.AddSpatialObject( box.GetPointer() )
> group.AddSpatialObject( gaussian.GetPointer() )
> 
> filter = itk.SpatialObjectToImageFilter[SOType, InternalImageType].New( 
> group, Size=[100,100], UseObjectValue=True )
> filter.Update() # required ?!
> 
> rescale = itk.RescaleIntensityImageFilter[InternalImageType, 
> OutputImageType].New( filter, 
> OutputMinimum=itk.NumericTraits[OutputPixelType].NonpositiveMin(), 
> OutputMaximum=itk.NumericTraits[OutputPixelType].max() )
> 
> itk.write(rescale, argv[1])
> 
> 
> --Gaëtan Lehmann
> Biologie du Développement et de la Reproduction
> INRA de Jouy-en-Josas (France)
> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
> http://voxel.jouy.inra.fr
>