Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!fs7.ece.cmu.edu!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uunet!hobbes!earth.armory.com!rstevew
From: rstevew@armory.com (Richard Steven Walz)
Subject: Re: Methods for sensing color electronically ???
Organization: The Armory
Date: Tue, 11 Oct 1994 09:28:01 GMT
Message-ID: <CxI4yr.87@armory.com>
References: <CxB2o8.7B@world.std.com> <373pb4$pdl@tabloid.amoco.com>
Sender: news@armory.com (Usenet News)
Nntp-Posting-Host: deepthought.armory.com
Lines: 34

In article <373pb4$pdl@tabloid.amoco.com>,
Jack Coats <jocoats@amoco.com> wrote:
>In article 7B@world.std.com, sja@world.std.com (Stuart J Adams) writes:
>>>How can I sense the color of an object electronically ???
>>>
>
>At robo-fest'94 in austin, TX, a guy had  robot that would follow people with red socks on,
>(or a red flag drug around a floor level).  My kid loved having it follow him
>around.  It had visible light sensors, 4 of them.  Two were covered with red
>celophane.  This was enough to enhibit most other frequencies of light.  It basically
>did the old go to where the light is brightest but just where the 'red' light is
>the brightest.
>
>It may be a cheap implementation, but it worked.  For an industrial
>strenght solution, you might want to put colored filters on several
>sensors, and calibrate them properly.  Especially with known color
>/light sources.  Or do color separation from a video camera.  One time
>in history, color video cameras were basically 3 B/W video tubes that
>had a theatrical jel over them (kind of like colored celephane :)
>and then combined the 3 video sources into one color signal later (all
>using the same sync signal of course!) 
>... JC
>E-mail: jocoats@amoco.com   Fax: 713/366-7570   Voice: 713/366-7120
>Personal: 73670.2476@compuserve.com -or- jcoats@hounix.org
---------------------------------------
Remember, JC, that all the beautiful Mars lander pictures from both Viking
landers on the surface of Mars were taken with a phototransistor, a mirror
that servoed, and a color wheel!!!! That'a all! It took HOURS and HOURS!!!
It did it one pixel at a time for one color, and then it moved the mirror
in one of two axes, both rotational. It was a simple pinhole camera to a
transmitter uplink!!!
-Steve Walz   rstevew@armory.com


