So far, I have only done tests in a controlled environment where I
can exclude interference from other light sources and ensure that only
uniform colours are in the sensor's field of view. For in less
well-defined circumstances, it should be possible to use modulated light
to eliminate daylight interference and add a lens or collimator to
look at only a small area.
Detecting colour sounds like it will be useful for all sorts of things.
Since everything that needs cutting is in the green/yellow part of
the spectrum, and most things that don't need cutting aren't, it should
provide a pretty good boundary detector that requires no training. But it
may also provide a means for placing simple signs in the environment
and prove useful in distinguishing paths that can be followed or
additional boundaries that should not be crossed.
Robin.
-- Copyright (C) 1996 R.M.O'Leary <robin nospam at acm.org> All rights reserved. For licence to copy, see http://dragon.swansea.linux.org.uk/mowbot/