Quick question: is Delauney triangulation the most appropriate registration technique, since its the only which uses local (as opposed to global) warping, which can account (somewhat) for pitch/yaw/roll issues which cannot be predicted globally? Are there other warping techniques available that can do a better job?
|