[Insight-users] Creating and Using Spatial Objects from Labeled Images

[Neal R. Harvey]

I would like to be able to start using more of the ITK functionality, 
with regard to registration.

In particular, I would like to register objects that I have segmented.

The result of the segmentation is a labeled connected component image
(created using  ConnectedComponentImageFilter followed by
RelabelComponentImageFilter).

Now, I would have thought that the next step would be to create a 
collection of spatial objects
and calculate some characteristics of these objects (such as 
orientation,  etc.) that might be useful
in the subsequent registering of the objects.

However, regarding this step of taking the labeled connected component 
image and creating the
spatial objects, I can't seem to find any good documentation for doing 
that. I would have thought
that this was quite common practice and there would be ample tools and 
associated documentation.
However, I have come up short so far in my searches.

I found a document entitled "Label Object Representation and 
Manipulation with ITK" by G. Lehmann,
which I thought was going to be helpful. But, it hasn't so far.

Any help with this would be much appreciated.