I need a routine that finds the best circle to fit a group of 3d points. I have the algorithms and am in the process of doing it myself, but time is short. Surely this problem has been solved before. It sure would be nice to have a routine that has already been tested by fire.
Here's the steps I'm presently trying to implement:
1) find the plane that best fits the 3d points. 2) project the points into the plane. 3) project the planar points from 3d to 2d. 4) find the circle that best fits the 2d points. 5) project the 2d points to the circle. 6) reverse everything to get back to 3d space. 7) display the center point, normal vector, diameter, variances, and standard deviation.
I've completed steps 1 - 3 above. Therefore, even a routine to find the best circle to fit a group of 2d points would be of great help. Hey, even a fit to cone routine.
If you can't help but think you know someone who can, let me know.
Respond by email, please. Thanks. -- Doug Taylor mailto:firstname.lastname@example.org (w) MailStop IW17C4 mailto:DUGnCON@aol.com (h) Intergraph Corp. (205) 730-6875 (w) Huntsville, AL 35894-0001 730-6708 (fax)
"Every parent homeschools, it's just a matter of degree."