Re: Warping surfaces to match control points

["Dick Jackson" <dick(at)d-jackson.com>]

"Dick Jackson" <dick@d-jackson.com> wrote in message
XO%07.9489$6N.15397@shaw-ty2">news:XO%07.9489$6N.15397@shaw-ty2...
>
> "Brad Nelson" <brad.nelson@nrc.ca> wrote in message
> 4e53fd15.0107050552.1442c0ea@posting.google.com">news:4e53fd15.0107050552.1442c0ea@posting.google.com...
> > I purchased a magnetic field grid which was generated from poor
> > navigation data. When we flew through the area again and measured the
> > magnetic field, many of the anomalies were displaced by km.  Does
> > anyone have an IDL procedure for warping an existing image (my grid
> > data) to a set of control points (my flight measurements). All of the
> > warping programs that I have seen rely on two images and a few "this
> > pixel should be over here" control points.  Nowhere have I found an
> > algorithm that takes many points and matches features.  Any
> > suggestions?
>
> I haven't used it before, but from the online help, I think WARP_TRI may
be
> just what you're looking for.

Forgive me for following up my own message, but I saw that I misread Brad's
question and was in touch with him directly. We concluded that it's really a
sort of registration problem he has, with a lot more flexibility in the
transform model than usually seen in mis-registration. Anyone with a better
idea than turning it into a big optimization problem?

Cheers,
--
-Dick

Dick Jackson / dick@d-jackson.com
D-Jackson Software Consulting / http://www.d-jackson.com
Calgary, Alberta, Canada