[vtkusers] Re: Implementing FFD as a VTK class

[Julia Schnabel]

Hi Gregg,

our lab ported the FFD transformation and registration software by Daniel 
Rueckert (IEEE-TMI 1999) to vtk. The transformation is indeed derived form 
vtkWarpTransform. Even though the registration does image intensity based 
optimization, you may be able to adapt it rather quickly to your own cost 
function (e.g. minimizing the distance between two pointsets). In fact, it 
probably contains some approximation routine based on linear and cubic B-splines 
for just that.

You can check it out on http://www.image-registration.com. Unfortunately, 
there's currently little to none support, as the main person responsible has now 
left, and I am on leave. There's a mailing list though which you might try (see 
link on web).

Good luck,
Julia

--
Julia A. Schnabel, Ph.D.
Computational Imaging Science Group    Tel: +44 (0)20 7955 2721
Division of Imaging Sciences                +44 (0)20 7955 4531 (secr.)
5th floor Thomas Guy House             Fax: +44 (0)20 7955 4532
Guy's Hospital, King's College London  mailto:julia.schnabel at kcl.ac.uk
London SE1 9RT, UK                     http://www-ipg.umds.ac.uk/j.schnabel/