Newsgroups: sci.image.processing
Path: cantaloupe.srv.cs.cmu.edu!europa.chnt.gtegsc.com!howland.reston.ans.net!news.nic.surfnet.nl!frontier.tno.nl!usenet
From: Jos Groote Schaarsberg <schaarsberg@tpd.tno.nl>
Subject: false color LUT
Message-ID: <1995Jun26.170529.18906@frontier.tno.nl>
Sender: usenet@frontier.tno.nl (Usenet Postmaster id)
Nntp-Posting-Host: tpdfp1.tpd.tno.nl
Organization: TNO - The Netherlands Organization for Applied Scientific Research
Date: Mon, 26 Jun 1995 17:05:29 GMT
Lines: 42

I'm often acquiring grey-value (0-255) images. This data can be a
normal (visual perceptible) 2D image but often contains non-standard
information e.g. temperature scan or (1 or 2 small bandwidth)
spectral characteristics. Within the images, we often want to compare
(or discern) the grey-value of features; these features can be (both)
in the same image but also can be in different images e.g. comparing
with a reference chart. The human perception cannot compare these
almost identical grey-values if these are not adjacent. For this
reason, I substitute the grey-values by colors using a false-color
LookUp Table.

I'm using a self-made false-color LUT, based on the (wavelength) 
spectrum, which converts each of the 256 grey-values to a unique 
combination of R-G-B. I'm not always pleased with the results of
this false-color LUT.

The applications are different from a label function i.e. one color
corresponds with e.g. the legend "sea" while another corresponds with
"desert". The applications compares quite well with the presentation
of a temperature scan i.e. "blue=cold" and "red=hot", although it is
not imperative that there is a "human-obviously" sequence of colors:
* With a LUT with 256 color shades, comparable colors should not be
conspicuous different
* With a limited number of color classes, all colors should be
good distinguishable from the other colors, while a low-contrast
image must not look like a funfair.
* A "human-obviously" sequence of colors is very helpfull when
lower/higher grey-value are important aspects.


Based on the technique of LookUp Tables wich converts a (grey-)
value to a color (R-G-B), I'm looking for additional false-color
tables: 
* LUT based on the spectrum but with selective spread of the
  magenta - blue - ... - red shades along the grey-values
* LUT based on a different, continuous order of color shades
* LUT based on reduction of the grey-values to a limited
  (e.g. 16) classes with each a color
* literature on the subject of composing false-color LUT's

Regards, Jos Groote Schaarsberg

