Article View: sci.math.num-analysis
Article #284Re: fitting circle to data points on small arc
From: pfennige
Date: Mon, 13 Apr 1992 16:57
Date: Mon, 13 Apr 1992 16:57
35 lines
1557 bytes
1557 bytes
Concerning fitting a circle to data points, recently
ritley@uimrl7.mrl.uiuc.edu
posted a message with the program FELLIPSE that fits an ellipse.
Instead of posting again the program FCIRCLE let me just mention that
a small modification to FELLIPSE fits circles by minimizing
sum( (X_i^2+Y_i^2 + d X_i + e Y_i + f)^2 )
The result is invariant by translation & rotation of the data points.
The following construction of the data matrix in FELLIPSE
> C Building the matrices A and B
> C
> DO 10 I = 1, N
> A(I,1) = X(I)**2-Y(I)**2
> A(I,2) = X(I)*Y(I)
> A(I,3) = X(I)
> A(I,4) = Y(I)
> A(I,5) = 1.
> 10 B(I) = -(X(I)**2+Y(I)**2)
> C
should be replaced by
> C Building the matrices A and B
> C
> DO 10 I = 1, N
> A(I,1) = X(I)
> A(I,2) = Y(I)
> A(I,3) = 1.
> 10 B(I) = -(X(I)**2+Y(I)**2)
and B(1), B(2), and B(3) are d, e, and f after the call to HFTI.
I can email FCIRCLE if this is not clear enough.
Daniel Pfenniger
Technical Details
Message-ID:
<1992Apr13.185746.1356@uni2a.unige.ch>Path:
arc1-ca.pugleaf.net!sparky!uunet!think.com!mips!darwin.sura.net!Sirius.dfn.de!chx400!ugun2b!ugun2a!pfennigeReferences:
<1992Apr9.003631.1350@magnus.acs.ohio-state.edu> <1992Apr10.171616.20698@ux1.cso.uiuc.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!think.com!mips!darwin.sura.net!Sirius.dfn.de!chx400!ugun2b!ugun2a!pfennige Newsgroups: sci.math.num-analysis Subject: Re: fitting circle to data points on small arc Message-ID: <1992Apr13.185746.1356@uni2a.unige.ch> From: pfennige@uni2a.unige.ch Date: 13 Apr 92 18:57:46 +0200 References: <1992Apr9.003631.1350@magnus.acs.ohio-state.edu> <1992Apr10.171616.20698@ux1.cso.uiuc.edu> Distribution: na Organization: University of Geneva, Switzerland Lines: 36