[SciPy-user] 2D and 3D moments of an image (Hu, Zernike, ...)

[Zachary Pincus]

Hi all,

My Grey-level coocurrence matrix code is indeed in ITK, and probably  
python-accessible via WrapITK. I can give some pointers to this if  
desired, but that's a pretty "heavy-weight" solution.

I think that having versions of these basic image feature  
calculations in a scikit or something would be very valuable. In the  
past, I've implemented Zernike moments and co-occurence matrix  
statistics in the context of ITK's C++ framework, and it really  
wasn't very much work. I would almost suggest that writing them from  
scratch as numpy operations might be easier than trying to deal with  
strange C-code. (And not slower -- I think that both can easily be  
cast in terms of ufuncs and whole-matrix operations, etc.)

I'm happy to talk at more length about the possibility of cobbling  
together such a scikit, if anyone's interested.

Zach


On Jan 22, 2008, at 12:24 PM, Stefan van der Walt wrote:

> Hi Matthieu
>
> On Tue, Jan 22, 2008 at 09:00:40AM +0100, Matthieu Brucher wrote:
>> On a side note, is there an efficient way to compute the  
>> cooccurence matrix of
>> an array ?
>
> You can find a ctypes implementation at
>
>   http://mentat.za.net/source/greycomatrix.tar.bz2
>
> This is a translated version of some C++ code I wrote a few years ago,
> so please check the results thoroughly (patches are welcome).
>
> Another person you may want to chat to is Zachary Pincus:
>
>   http://public.kitware.com/pipermail/insight-users/2004-June/ 
> 009190.html
>
> who also hangs around on this list.
>
> Regards
> Stéfan
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user