home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!destroyer!ubc-cs!uw-beaver!news.u.washington.edu!sumax!thebes!amc-gw!kenb
- From: kenb@amc.com (Ken Birdwell)
- Newsgroups: comp.graphics
- Subject: fitting to a NURBS surface
- Message-ID: <1992Jul23.035836.5217@amc.com>
- Date: 23 Jul 92 03:58:36 GMT
- Distribution: na
- Organization: Applied Microsystems, Redmond, WA
- Lines: 23
-
-
- I'm looking for some code (or at least a good reference) that will fit a NURBS
- surface to a bunch of 3D points (ie. X,Y,Z data). I've orginized the points
- so that they all have reasonable U,V coordinates (rangeing from 0 to 1), so
- everything is ready for spline surface fitting. I'm currently fitting a
- parametric polynomial surface, and it works fine, but I would much rather
- generate a NURBS surface.
-
- I have a lot of points (around 10,000), so I need an algorithm that
- generates far fewer nodes than the number of points. I know this will only
- appoximate the surface, but since the data's fairly noisy, this turn out to
- be a good thing.
-
- I've found a lot of papers that discuss specific things like "blossoming"
- and other node selection methods at a fairly sophisticated level, but I need
- something that talks about fitting NURBS at a much more basic level.
- Something along the lines of pseudo-code would be great, I'm not much of a
- mathematician. The only papers I've been able to find at that basic of a
- level only deal with lines, and I'm at a loss as to how to extend their
- descriptions into surfaces. Any help would be greatly appreciated.
-
-
- --
-