[Insight-developers] code production line & registration refactoring

[Cory Quammen]

Brian,

I haven't had a chance to look at your registration refactoring
proposal, but I support the separation of the image transformation
from the metric computation. This should enable other uses of the
optimizers and metrics in ITK, such as fitting a parametric image
model (e.g., a model of a widefield fluorescence microscope
point-spread function) to an image, in a way that is less kludgy than
what I am doing now.

I'm interested in attending a telecon on the refactoring of the
registration classes if you schedule one.

Cory

On Fri, Aug 27, 2010 at 10:54 AM, brian avants <stnava at gmail.com> wrote:
> just a quick email to see that i understand the dev process
> that we might want to use for registration refactoring:
>
> 1.  check out the current development version of itk
>
> 2.  create a topic branch , e.g. registration
>
> 3.  develop unit tests using the google unit test style ?
>
> 4.  write some code --- local commits etc --- make sure tests pass.
>
> 5.  uncrustify , kwstyle , review the code , document , etc.
>
> 6.  push to the git topic branch --- or is this where gerrit comes in?
>
> is this correct?  i need to be able to communicate a cohesive dev
> procedure to the rest of the development team here and i'm having a
> hard time finding a single doc that covers the whole system ....
>
> re: registration refactoring --- see the proposed plan in the shared
> google docs :
>
> ITKv4 =>  v4 Proposals
>
> we are currently working on refactoring the linear registration metric classes
> such that the transformation of the object (e.g. image) is separated from the
> metric computation.    this just means that transformation of the
> object happens
> outside of the metric ( currently, transformation of the objects happens inside
> the metrics ).    there are a number of simplifications that this
> strategy will allow ,
> including enabling a generalized (e.g. sparse) sampling strategy for
> all the metrics,
> symmetric / robust registration ( e.g.
> http://reuter.mit.edu/index.php/publications/ ) , masked registration
> and it can probably be implemented in a backwards compatible way ...
>
> discussing this has been a tcon topic for some time but we havent
> gotten to it probably
> because there have been so many other lower level bases to cover.
>
> would be happy to schedule a specific tcon time or a separate conf
> call if people are
> interested.
>
> brian
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://kitware.com/products/protraining.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
>



-- 
Cory Quammen
Center for Computer Integrated Systems for Microscopy and Manipulation (CISMM)
Department of Computer Science
University of North Carolina at Chapel Hill
http://www.cs.unc.edu/~cquammen