Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!lll-winken!uunet!zaphod.mps.ohio-state.edu!sdd.hp.com!ucsd!sdcc6!jclark
From: jclark@sdcc6.ucsd.edu (John Clark)
Newsgroups: comp.multimedia
Subject: Re: affine transform
Keywords: compression, affine transform
Message-ID: <16177@sdcc6.ucsd.edu>
Date: 30 Jan 91 19:36:30 GMT
References: <1991Jan22.220948.12523@pandora.matrox.com>
Organization: University of California, San Diego
Lines: 19

In article <1991Jan22.220948.12523@pandora.matrox.com> ltran@pluton.matrox.com (Linh TRAN) writes:
+
+	I heard recently about a new class of transform (at
+least new for image compression purpose). My questions are:

It sounds like the Iterated Function Systems. There is a book by an
author named Beardsley(sp). I'm at work now, but will post the full
title and ISBN later today(1/30/90).

Briefly, an IFS is a set of coefficients for an affine X-form and an
interation sequence which will 'recreate' an area of an image. The
number of coefficients is small relative to the area of the image
being recreated, hence compression. The system does not create exact
fidelity but recreates the 'visual' impression of the original to
varying degrees of fidelity. 
-- 

John Clark
jclark@ucsd.edu