[Insight-developers] Re: integrating the new algorithms for existing filters in ITK (preparing ITK 3.2)

[Gaetan Lehmann]

On Tue, 12 Dec 2006 00:16:03 +0100, Luis Ibanez <luis.ibanez at kitware.com>  
wrote:

> If you have a chance, please take a look at list that Julien put
> together and list the papers that you see have already enough
> reviews and that, in your opinion, will be very useful if put
> in the toolkit. It will be great if you can participate on the
> tcon this Friday, but if not, at least, we will appreciate to
> have your list of suggested papers to integrate in ITK 3.2.
>

I'll try to be there.
In any case, here is my list:

Very useful, clean, well tested, and well known (by me) contributions:

   The watershed transform in ITK - discussion and new developments
   MinimaImpositionImageFilter
   Image projections along an axis
   Label overlay
   Finding regional extrema - methods and performance
   Binary morphological closing and opening image filters
   FFTShift

Well, that's Richard's contributions and mine, but that's the one I know  
the best, so I'm pretty sure about their quality.



Very useful, clean, well tested, and well known (by me) existing code  
enhancement:

   Improving performance of morphological reconstruction
   Optimization of connected component labelling
   Improving features and performance of binary erode and dilate filters

Still Richard's contributions and one of mine - we must do a very good job  
:-)



The contributions I have done and shouldn't be included (yet) - no need to  
spend time on these ones:

   Slice by slice filtering with ITK - still some work to do
   Robust Automatic Threshold Selection - need some reviews
   Kappa Sigma Clipping - lot of non published changes already done, need  
more explanations, need some reviews
   Consolidated morphology - not fully ready



Contributions I would argue against their integration, despite their  
quality:

   Parallel 3D Exact Signed Euclidean Distance Transform, for its 3D  
limitation
   Digital Topology, because the connectivity framework need some extra  
work to support iterators, and to remove the connectivity option from the  
template parameter



For Julien: The following ones are already in the toolkit:

   itkConnectedComponentFunctorImageFilter
   N-D Linear Time Exact Signed Euclidean Distance Transform


Gaetan

-- 
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