If Karl's code works ok then it would be nice to make a high level interface 
that does the BB optimization and returns a mask. That could be added to some 
appropriate lib part of the dist (Image2D?).

Christian

On 11/09/2010, at 9:22 AM, Matthew Kenworthy wrote:

> 
> I benchmarked the four programs with a 1000 x 1000 input piddle and
> small convex and concave polygons (really small, only 5-6 points
> each).
> 
> 
> My solution is going to choke for lots of points, since it generates a [N, 
> dims(image)] size mask. I was looking into the winding soution, but Karl's 
> code is an optimised version of that and so I'd suggest you stick with that.
> 
> At the end of the day, the main thing is...does it do the job you want 
> correctly, and does it do it fast enough? I love working with PDL, but if 
> your code is good enough  for the problem, go for it!
> 
> OTOH, if you're using it as a learning experience, you can never get enough 
> examples :)
> 
> Matt
>   
> _______________________________________________
> Perldl mailing list
> [email protected]
> http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

--
Christian Soeller PhD   Dept. of Physiology  +64 9 3737599 x82770
University of Auckland  Auckland, New Zealand  fax +64 9 3737499


_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to