[Insight-developers] New ITK data objects

[Luis Ibanez]

Hi John,

I would like to insist in that implementing a ND
version is no harder than implementing a 2D one.

The chain code class will still have just an Index
and a std::string as ivars, no matter the dimension
of the space.



   Luis


---------------------------------
John M. Galeotti wrote:
> I am not opposed to the idea of a general "path" class, although I would
> like to solicit suggestions as to what belongs in the general API. Are 
> Add/GetOrigin, Add/GetStep, and GetNumSteps sufficient? I would leave it 
> to someone else to implement the minimum cost path calculation, etc.
> 
> My ultimate vision for the parametric function hierarchy would be to
> have generic ND versions implemented and a special, efficient 2D
> sub-hierarcy implemented as well.  However, since I only need the 2D
> sub-hierarchy for my work, I currently do not intend to implement the ND 
> versions (at least not yet; anyone else interested?).
>