[Insight-developers] Changing wrapped image types

Miller, James V (Research) millerjv at crd . ge . com
Fri, 11 Jul 2003 14:43:47 -0400


Luis,

I already have modified the examples thanks to an Emacs macro.

Are there any other wrapping things lying around?  Under Testing?
InsightApplications?

Jim

> -----Original Message-----
> From: Luis Ibanez [mailto:luis . ibanez at kitware . com]
> Sent: Friday, July 11, 2003 2:40 PM
> To: Miller, James V (Research)
> Cc: Insight-developers (E-mail)
> Subject: Re: [Insight-developers] Changing wrapped image types
> 
> 
> 
> Hi Jim,
> 
> The Python and Tcl examples in Insight/Examples will have
> to be modified a bit. (mostly for changing US2 -> S2...)
> 
> This shouldn't be too much of a problem.
> 
> I'll be happy to make those changes, once you are done
> rearranging the filter types to be wrapped.
> 
> 
>    Luis
> 
> 
> -----------------------------
> Miller, James V (Research) wrote:
> > I'd like to check in some changes to the type of 
> images/filters that are 
> > wrapped by Swig.  We discussed earlier this week that 
> wrapping float and 
> > short would cover most use cases.  We also need to wrap 
> unsigned long 
> > since that is what we typically use for label maps. To 
> verify that we 
> > can wrap non-native pixel types, I am also wrapping 
> floating point rgb 
> > images. We'll probably also need to wrap a covariant vector 
> image type.
> >  
> > Users will always have the option of wrapping more image 
> types and more 
> > filter instantiation combinations.  I'd just like to settle 
> in on what 
> > default types to start using.
> >  
> > Bill mentioned that people are using the Swig wrapping.  If 
> I make these 
> > changes, then some of the current filter instantiations will be 
> > removed/replaced.  For instance, the US2US2 filters will 
> become S2S2 
> > filters.
> >  
> > Will this be disruptive to the people that are currently 
> using the Swig 
> > wrappers?  I can hold off until next week if that is better 
> for developers.
> >  
> > Jim
> 
> 
>