Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!howland.reston.ans.net!news.sprintlink.net!alfa02.medio.net!netnews.nwnet.net!serval.net.wsu.edu!unicorn.it.wsu.edu!a0014246
From: a0014246@unicorn.it.wsu.edu (mark fuller)
Subject: Re: Question: Range finder in photocameras. How?
Message-ID: <D4pCCo.5M1@serval.net.wsu.edu>
Sender: news@serval.net.wsu.edu (News)
Organization: Washington State University
X-Newsreader: TIN [version 1.2 PL2]
References: <KARL.17.0014E4D5@innet.be>
Date: Tue, 28 Feb 1995 08:46:00 GMT
Lines: 16

Karl Lambrechts (KARL@innet.be) wrote:
: I put this question already once. I only got 1 reply ( however interesting, 
: but very complicated for sure)

: This system must be very simple because you can find it in rather cheap 
: photocameras. I studied mine's and you can clearly see two lenzes, completely 
: seperated from the other optics.

Fairly simple, really.  One lens is used to focus the output of an infrared light emitting diode to a narrow beam.  Behind the other lens, which is usually as far from the first lens as design allows is a small linear array of charge coupled devices on a single chip.  This chip can detect the reflection of the beam from the target to be focused on.  The way it knows the range is since the lenses are x distance apart, for a given distance to target is at a slight angle from straight ahead from the detector.

  The ccd can measure this offset angle and plug it into a formula (tangent of angle = (distance between lenses divided by distance to target)).  Since ccd's are digital devices, the output is in steps, usually 128 or 256 of them.

Hope this answers your question.

Mark Fuller
TANSTAAFL
