Path: utzoo!mnetor!uunet!mcvax!unido!ecrcvax!johng
From: johng@ecrcvax.UUCP (John Gregor)
Newsgroups: comp.graphics
Subject: Re: Fractals!
Message-ID: <526@ecrcvax.UUCP>
Date: 29 Apr 88 10:30:13 GMT
References: <811@unioncs.UUCP> <1776@uhccux.UUCP> <5595@pyr.gatech.EDU> <931@stracs.cs.strath.ac.uk>
Reply-To: johng@ecrcvax.UUCP (John Gregor)
Organization: ECRC, Munich 81, West Germany
Lines: 26
Keywords: Fractals, Mandelbrot, Julia, Chaos, Iteration in the Complex Plane

In article <931@stracs.cs.strath.ac.uk> rk@cs.strath.ac.uk writes:
>In article <5595@pyr.gatech.EDU> paulm@pyr.gatech.EDU (PAUL MILLER) writes:
>>   The gist of the article involves representing ANY image by its IFS codes, 
>>thereby achieving significant image compression ratios (> 10,000 : 1).
>I am impressed by the ratio of 10,000:1.  Can it be true?
>
>PKARC.  It often achieves compression ratios of 3:1, but I should certainly
>like to see 10,000:1.  Let's see :- a picture 512 X 512 pixels using one
>byte per pixel occupies (uncompressed) 262144 bytes.  Using the IFS system
>mentioned this is compressed to just 27 bytes! Fantastic!

Unfortunately, the search space is ~ 2 ^ (27*8)  or  2 ^ 216.  And even
then the compression will be destructive (i.e. the uncompressed picture
won't be the same as the original.  I've only seen it work on demonstration
pictures that looked very fractalish to begin with.  So I can "compress"
the Mandelbrot set into a handfull of bits, big deal.  It still takes a
couple of hours to get at the areas of the set I want to see.

(* inews laxative *)
(* inews laxative *)
(* inews laxative *)
-- 
pqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpqpq
bdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbdbd

John Gregor                                     johng%ecrcvax.UUCP@germany.CSNET