[Insight-users] edge preserving smoothing filter

[Miller, James V (Research)]

GradientAnisotropicDiffusionImageFilter is the standard approach. The other
variants like CurvatureAnisotropicDiffusionImageFilter produce images that 
are more "piecewise" constant which can be useful for extracting objects of 
interest.

I have not looked at the background of the CurvatureImageFilter. So I cannot

comment on its relative merits to the above diffusion operators.

The BilateralImageFilter does not have the characteristic ripple (caused by
sharp discontinuities) as the diffusion operators.  This is appealing for
some
applications.  But the BilateralImageFilter is very slow in comparison to
the
diffusion methods.

Hopefully we'll have a version of Mean Shift image filtering soon.  It
should behave
similar to the bilateral filter.

We recently added some grayscale morphological operators based on geodesic 
erosion and dilation.  These are also quite useful for removing noise. 

GrayscaleGeodesicErodeImageFilter
GrayscaleGeodesicDilateImageFilter
HMaximaImageFilter
HMinimaImageFilter
HConvexImageFilter
HConcaveImageFilter




> -----Original Message-----
> From: salah [mailto:salah at gris . uni-tuebingen . de]
> Sent: Wednesday, July 02, 2003 10:03 AM
> To: ITK Users (E-Mail)
> Subject: [Insight-users] edge preserving smoothing filter
> 
> 
> hello all,
> 
> depending on your experiments, what is the best edge-
> preserving smoothing filter to use with CT medical images.
> Speed/quality ??
> 
> Thanks,
> Zein
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk . org
> http://www . itk . org/mailman/listinfo/insight-users
>