NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/comp/graphics/12126

Internet Message Format | 1992-11-19 | 2.7 KB

Path: sparky!uunet!munnari.oz.au!uniwa!madvax.uwa.oz.au!watson
From: [e-mail] (David Watson)
Newsgroups: comp.graphics
Subject: Re: Surface Representation Using Patches.
Date: 20 Nov 1992 03:30:33 GMT
Organization: Maths Dept UWA
Lines: 39
Distribution: world
Message-ID: <[e-mail]>
References: <[e-mail]>
NNTP-Posting-Host: xanthorrhoea.maths.uwa.oz.au

In article <[e-mail]>, [e-mail] (Rached Zantout) writes:

|> I am trying to survey techniques on how to represent a surface using patches.
|> What I will be looking for is methods that people have developed that will
|> produce a patch representation of a digitized surface. Also I would be looking
|> at the goals behind those methods and the things they could not acheive. My
|> final goal is to determine whether there exists room for a "better" way or
|> whether the literature has already the "best" way to do that.

The literature contains many ways of generating patchwise surfaces for both
gridded data and scattered data.  The major trick is to make the slopes of the
patches match, and this is especially difficult if the patches are irregular.

|> The problem I think arises when a surface is to be represented by more than one
|> patch. We will have to determine which points contribute to which patch and
|> what is the size of each patch. Whether we should allow different sized patches
|> is also a question to be solved.

Yes, subset selection is extremely difficult for scattered data.  Using either a
fixed number of data or all the data within a fixed area can lead to excess or
insufficient data for a given patch if data density is uneven or anisotropic.
Second order natural neighbor subsets are the only subsets, that I am aware of,
which will overcome this problem.  However, given such subsets, one cannot use
any of the usual spline patches because of slope discontinuities at the edges
of the patches; natural neighbor interpolation can be used to generate a surface
with a continuous slope.

|> If anybody has information or pointers to where can I start my literature
|> review, or better yet if I can find a survey paper that already did the survey
|> work, I will be very grateful.

For some 600 references to contouring papers, see
Watson, D.F., 1992, Contouring - a guide to the analysis and display of spatial
data: Pergamon, ISBN 0 08 040286 0 (H)
--
Dave Watson                          Internet: [e-mail]
Department of Mathematics
The University of Western Australia               Tel: [phone]
Nedlands, WA 6009  Australia.                     FAX: [phone]