Subject:
|
Re: GPS Reality?
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Tue, 12 Apr 2005 00:57:18 GMT
|
Original-From:
|
Steve Baker <sjbaker1@+stopspam+airmail.net>
|
Viewed:
|
1153 times
|
| |
|
|
danny staple wrote:
> A solution I suggested on the DPRG list was that if radio telemetry is
> being sent to and from the robot, then to have each robot with a
> clearly identifiable dot- rather like a police car - on the top. You
> could then have a camera covering the arena, and a desktop machine
> with visual processing software to work out the relative position of
> the dot, then communicate this back to the bot with the telemetry.
> However - this would only work in a very small space.
With a cheap digital camera (say 800x600 resolution) with a 2.5meter
ceiling (around 8') looking straight down through a typical 60 degree
field of view lens, you're limited to an area about 2m wide by 1.5m
wide...less if you are working on a table-top rather than on the floor.
You could of course use more cameras.
If you can get a wider angle lens or a higher ceiling, you could do
a larger area...but even over a two meter table-top, your resolution
(with an 800x600 pixel unit) is a quarter centimeter - that's pretty
good - and better cameras (up to 3000x2000 pixels or so) are available
if you have deep pockets.
> Also - it would limit the choice of colours, as Lego colours tend to
> be very saturated colours, which are what this technique normally uses
> to find stuff. We are back to visual processing now - but at least it
> is now taken off the bots.
The way to avoid the colour problem is to put a retro-reflector on the
robot. The camera can then take a photo with the flash on. The
retro-reflector will appear so bright that everything else will be
almost black.
It's a reasonable solution - but it's no good if the robot has to be
able to drive under things.
It's quite possible to come up with good ideas for heavily restricted
situations...it's the general solution that's tough.
Think robotic vacuum cleaner. Imagine bouncing down staircases, running
around under a table between a maze of chairlegs. Think about the family
dog who sleeps wherever it feels like it and moves around at other times.
---------------------------- Steve Baker -------------------------
HomeEmail: <sjbaker1@airmail.net> WorkEmail: <sjbaker@link.com>
HomePage : http://www.sjbaker.org
Projects : http://plib.sf.net http://tuxaqfh.sf.net
http://tuxkart.sf.net http://prettypoly.sf.net
-----BEGIN GEEK CODE BLOCK-----
GCS d-- s:+ a+ C++++$ UL+++$ P--- L++++$ E--- W+++ N o+ K? w--- !O M-
V-- PS++ PE- Y-- PGP-- t+ 5 X R+++ tv b++ DI++ D G+ e++ h--(-) r+++ y++++
-----END GEEK CODE BLOCK-----
|
|
Message is in Reply To:
 | | Re: GPS Reality?
|
| A solution I suggested on the DPRG list was that if radio telemetry is being sent to and from the robot, then to have each robot with a clearly identifiable dot- rather like a police car - on the top. You could then have a camera covering the arena, (...) (20 years ago, 11-Apr-05, to lugnet.robotics)
|
6 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
