> > These lines are used by the 'morph-to-fit' feature:
> > C i0 x0 y0 X0 Y0
> > means: After remapping image 0 (i0), take the point with coordinates
> > x/y, and move
> > it to position X/Y. This will only happen, if you specified 'o'
> > in the corresponding image line. You can manually edit these 'C'-lines
> > to change the positions.
> 
> This sounds interesting, is there any documentation about this feature? What
> does the morphing do?
> 

The tutorial example from my website uses this feature. I think
there is some mentioning in the script-file but no serious
documentation. We had a thread on this list some months ago.

The morphing algorithm triangulates the panoramic image.
Each control point becomes the corner of a triangle. The
triangle mesh is then distorted according to the 'C'-lines.
This guarantees that the control points match exactly. This feature
can be used in hopeless cases, eg parallax errors or
weird distortions.

I should add that _all_ transformations (remapping + 
whatever correction is specified + morphing) are performed 
in a single step requiring just one sampling of the 
original source image.

The PTMorpher application of the PanoTools package
provides access to the morphing functions without
any of the panoramic stuff.

Regards

Helmut Dersch