Summary on Distortion Transforms

[anthony]

FYI...

As part of an on going email discussion on Perspective transforms, I have tryed to summarize exactly what a 'Distrotion transform' is and how it is 'generally', (as apposed to actually) implemented.

http://www.cit.gu.edu.au/~anthony/graph ... s/#summary

I would like your view on this summary.

With this summary, I can invisage a more 'generalized' and object orientated transformation algorithm, that can be used for many of the existing distortion transformations, including distortion maps (positional and displacement), affine transformations, rotations, shears, and even perspective, that may be a lot faster given the current very complex multi-stage algorithms being used.

Comments?