Re: A question on algebraic circle fitting
- From: "Robert Israel" <israel@xxxxxxxxxxx>
- Date: 24 Oct 2006 22:57:05 -0700
Han de Bruijn wrote:
Thomas Mautsch wrote:
Given a finite number of points in the plane,
(x1,y1), (x2,y2), ..., (xn,yn),
the "algebraic way" to fit a circle to these points
is to minimize the sum
sum( (xi^2 + yi^2 + 2 D xi + 2 E yi + F)^2 , i = 1..n )
over the variables D,E,F.
Is it correct that the resulting minimizers
will satisfy the condition
D^2 + E^2 >= F ?
Under what conditions will minimizers D,E,F exist? -
A necessary condition is that not all points (xi,yi) lie on
a common line in case n >= 3. Is this condition also sufficient,
or what other conditions are there?
http://huizen.dto.tudelft.nl/deBruijn/programs/delphi.htm#BFC
Read the paragraph "Method by RI" (: Robert Israel). Yes, it is correct
that the resulting minimizers will satisfy the condition D^2 + E^2 >= F
as far as I can see.
I find no "Method by RI" on that page. Perhaps you mean
http://hdebruijn.soo.dto.tudelft.nl/jaar2006/kromming.pdf
Robert Israel israel@xxxxxxxxxxx
Department of Mathematics http://www.math.ubc.ca/~israel
University of British Columbia Vancouver, BC, Canada
.
- References:
- A question on algebraic circle fitting
- From: Thomas Mautsch
- Re: A question on algebraic circle fitting
- From: Han de Bruijn
- A question on algebraic circle fitting
- Prev by Date: Re: An uncountable countable set
- Next by Date: Re: Find the largest integer a or b without comparisons??
- Previous by thread: Re: A question on algebraic circle fitting
- Next by thread: Re: A question on algebraic circle fitting
- Index(es):
Relevant Pages
|
