Path: utzoo!attcan!utgpu!watmath!maytag!watstat!dmurdoch
From: dmurdoch@watstat.waterloo.edu (Duncan Murdoch)
Newsgroups: comp.graphics
Subject: Re: How to map 24-bit RGB to 256 co
Message-ID: <443@maytag.waterloo.edu>
Date: 5 Sep 89 14:40:24 GMT
References: <7772@cbmvax.UUCP> <1109@krafla.rhi.hi.is>
Sender: daemon@maytag.waterloo.edu
Reply-To: dmurdoch@watstat.waterloo.edu (Duncan Murdoch)
Organization: U. of Waterloo, Ontario
Lines: 14

In article <1109@krafla.rhi.hi.is> pierre@rhi.hi.is (Kjartan Pierre Emilsson) writes:
>A year ago or so, there was a similar discussion and there a method originally
>devised by Paul Heckbert was introduced, called Color Cube Compression. 
>I don't have the original posting but the method went something like that:

The description you gave of the algorithm (which someone else said you 
should have called the Median Cut algorithm) sounds a lot like a statistical
technique called regression trees.  The main difference is that you chose
to divide your rectangles at the median of the longest axis, while the 
regression tree algorithm looks for the "most discriminating" place to cut.
I think the standard reference is Classification and Regression Trees;
L. Breiman, J.H. Friedman, R.A. Olshen, C.J. Stone; Wadsworth, 1984.

Duncan Murdoch