Newsgroups: sci.image.processing
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!zombie.ncsc.mil!news.duke.edu!convex!cs.utexas.edu!news.sprintlink.net!uunet!ftpbox!news.acns.nwu.edu!news.eecs.nwu.edu!k9ape
From: k9ape@eecs.nwu.edu (Sheldon L. Epstein)
Subject: Re: DETECTING EYE-POSITION
Message-ID: <D59Aq7.4uJ@eecs.nwu.edu>
Sender: usenet@eecs.nwu.edu
Organization: EECS Department, Northwestern University
References: <Pine.HPP.3.91.950227030933.3514A-100000@mandinga.cip.informatik.uni-muenchen.de> <3ji55k$2mg@ddi2.digital.net>
Date: Sat, 11 Mar 1995 03:22:54 GMT
Lines: 38

In article <3ji55k$2mg@ddi2.digital.net>,
Nathan A. Harris <nharris@dba-sys.com> wrote:
>In article 
><Pine.HPP.3.91.950227030933.3514A-100000@mandinga.cip.informatik.uni-muen
>chen.de>, greiner@mandinga.cip.informatik.uni-muenchen.de says...
>>
>>Has anybode heard about methods to detect the position of eyes?
>>I am building up an medical experiment where I have to know where the 
>>eyes are. I want to film the face by video and need a Image recognition 
>>to find the eye position.
>
>
>There is an image processing device which tracks the movement of a
>subject's eyes.  However, you must initially identify where the eye
>is and then it will follow it's movement.  The output of this eye
>tracker can be in video pixels or it can output where the subject is
>looking.
>
>Would this achive your goals?  Send me email for more info.
>
>Nathan
>nharris@dba-sys.com
>

Hello Nathan,

The problem was solved years ago and answers have been described
in numerous publications in Bio-Medical Engineering.

Basically, flood the face with near-infrared illuminatiion.  Then
get a good black & white video camera from which you can extract
its IR-blocking filter (which is usually located behind its C-Mount
threads).  Select an IR-friendly lens and go for it,  The rest
will become obvious.

			Shel Epstein, k9ape@eecs.nwu.edu
			Chief Engineer
			Epstein Associates
