optimization - Need a suggestion on curve processing (curve-fitting, interpolation) -
i have set of points form curve, see picture. curve consists of straight , curved sections , not know start or end. in presented image know @ to the lowest degree 5 pieces: 2 straight sections, 1 section constant radius, 2 other types of curves. want able recognise sections, separate them each other , work them separate curves.
update.
in sentiment splines not work in way want, , not certainly besier. thinking non-linear to the lowest degree squares not sure if suits case. if able separate parts, can utilize linear to the lowest degree squares straights , non-linear other parts. otherwise, need universal method work types of curves: 1, 2 , 3 power.
please share thoughts. give thanks you.
for bent tube suggest multidimensional spline fitting. tube not composition of of straight , curved sections, curves smoothly way.
http://en.wikipedia.org/wiki/spline_interpolation
this starting point if want read further. taking guess suppose not experienced splines, if set out dataset in table form me or other reader interpolate given info 3d spline curve.
optimization interpolation curve-fitting
No comments:
Post a Comment