Path: utzoo!mnetor!uunet!husc6!think!ames!pasteur!ucbvax!decvax!decwrl!labrea!csli!rustcat
From: rustcat@csli.STANFORD.EDU (Vallury Prabhakar)
Newsgroups: comp.graphics
Subject: Re: Algorithm wanted: Circle enclosing points
Message-ID: <3297@csli.STANFORD.EDU>
Date: 4 Apr 88 08:45:20 GMT
References: <wWIfSMy00Xo3A5u1dC@andrew.cmu.edu> <496@etn-rad.UUCP> <3229@phri.UUCP>
Reply-To: rustcat@csli.UUCP (Vallury Prabhakar)
Organization: Yonder, the Apocalyptic Horizons
Lines: 19

This is pretty trivial.  Pick any point in the set as the centre.  Compute
the largest distance from the remaining points.  That will give you the 
radius of the circle (after adding an appropriate factor for enclosure as
opposed to lying on the circumference, of course).

Question:

Has anyone attempted to solve for curve intersections where the curves are
represented in the parametric cubic form such as by Hermite cubics? If so, 
I'd like to hear about the equation solver that's used.  The above method
gives rise to a system of 2 non-linear equations in the 2 parametric unknowns.
Given the restrictions on the parameters (0 <= u,v <= 1) and the fact
that the number of intersection points cannot exceed two, I was wondering
if there was some special technique of equation solving that could use these
pieces of information.  

Thanks in advance for any suggestions.

						-- Vallury Prabhakar
						-- rustcat@cnc-sun.stanford.edu