Re: Discrete Curvature and Torsion for 3D-poly line
- From: "Roger Stafford" <ellieandrogerxyzzy@xxxxxxxxxxxxxxxxxxxxxx>
- Date: Wed, 16 Apr 2008 21:51:02 +0000 (UTC)
"Nodame Han" <tmeister@xxxxxxxxx> wrote in message <fu5pd4$2b1
$1@xxxxxxxxxxxxxxxxxx>...
Hi, I'm trying to calculate torsion and curvature of a 3D------------
Curve, and then I'll compare different curves with each
other by using them.
Actually, it is composed of 3D-poly Lines connected to each
other by 3-d points (x,y,z). So, I don't think it's
necessary to calculate spline or interpolation (because
every curve is composed by same way).
It would be possible to get the profiles of torsion and
curvature by calculating kind of discrete version of them
at the turning points.
But I don't know how to...
Is there any good methods or reference?
Thanks.
As I see your problem, it will take a minimum of three neighboring points to
determine an approximate osculating plane, an approximate curvature
therein, and the binormal direction to it. At least one more point would be
required to determine the rate of change of this binormal to get an
approximate torsion. You will therefore need to estimate your curvature and
torsion using neighborhoods of at least four points and preferably many
more.
My octogenarian memory of the old differential geometry days is pretty hazy
by now, however. It would be better if you could find this problem already
worked out somewhere on the internet. Otherwise, I would have to put the
little grey cells through an awful lot of work.
Roger Stafford
.
- References:
- Discrete Curvature and Torsion for 3D-poly line
- From: Nodame Han
- Discrete Curvature and Torsion for 3D-poly line
- Prev by Date: Multiple COM ports
- Next by Date: Re: Discrete Curvature and Torsion for 3D-poly line
- Previous by thread: Discrete Curvature and Torsion for 3D-poly line
- Next by thread: Re: Discrete Curvature and Torsion for 3D-poly line
- Index(es):
Relevant Pages
|