Newsgroups: comp.robotics
Path: brunix!sgiblab!swrinde!emory!sol.ctr.columbia.edu!The-Star.honeywell.com!umn.edu!csus.edu!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: Laser Range Finding
Message-ID: <1993Feb6.151229.13798@netcom.com>
Organization: Netcom - Online Communication Services  (408 241-9760 guest) 
References: <1993Feb5.175607.24099@ncsu.edu>
Date: Sat, 6 Feb 1993 15:12:29 GMT
Lines: 37

adhudson@eos.ncsu.edu (ALAN D HUDSON) writes:
>One idea I had, which I admit I have no real clue as to if it will work is this:

>Using a laser diode with a intensity measuring device I would swing the 
>diode and collector around in a 360 degree arc.  The idea is this:  take a 
>background reading on the sensor, pulse the diode and take another 
>reading, the intensity of the return beam should give me distance.  I 
>"know" this will work for IR, but I think it won't give me the range/
>accuracy I need.  The only problems I see are tese:  there is not enough 
>spread on a laser to measre with a sensor not perfectly aligned to the 
>return beam?

       That's a reasonable idea.  

       If you have two receive sensors, placed at different distances,
you can use the ratio of the intensities received at both of them to
determine range.  This works independent of the reflectance of the
target.  This is a very effective technique for short-range sensors,
as used in hand-grasping control, but it's difficult to use it over
a range of several meters.  Without this, though, you won't really get
a range value, just a reflectance value biased by distance.

       To estimate intensity, you can assume all the light hits the target,
and that the target reflects some percentage of the light, spreading 
out the light over a hemisphere.  (This is
a good model for, say, painted surfaces.)  The surface area of the 
hemisphere is proportional to the square of the range, so the received
intensity is proportional to (1/(d^2))*r, where d is the distance (range)
and r is the reflectance of the target surface.

       For an example of an image produced by a sensor of this type,
see "Combining Artificial Neural Networks and Symbolic Processing for
Autonomous Robot Guidance", by Pomerleau et al, in Proc. Image Understanding
Workshop, January 1992, Morgan Kaufman, ISBN 1-55860-244-5.  Look for the
laser reflectance image.

					John Nagle
