Subject:
|
Re: proximity detection with laser?
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 5 Jan 2001 20:56:31 GMT
|
Original-From:
|
Hao-yang Wang <hywang@pobox&Spamless&.com>
|
Viewed:
|
766 times
|
| |
|
|
> You need either two camera's and one laser, or you need two lasers and one
> camera. The system I prefer (and built many moons ago) is a dual laser
> system and the two beams are collimated such that they cross (ie form a
> single dot) at a well defined distance. Then use the camera to measure the
> distance between the two dots (and it's rate of chance) to deduce distance
> by simple trig. I think you'll find this more accurate and easier to
> program than the two camera systems (ie image analysis/comparison isn't
> the best understood problem).
When I said one dot is enough, I was thinking about those IR-based
distance measuring sensors from Sharp,
(See <http://www.sharp.co.jp/ecg/opto/products/f-osd.html>.), but you are
right, two dots simplify things a lot.
In my current project, I mount a wireless camera on top of the mobile
robot, to look for barcode tags as the landmarks. I use CyberMaster
instead of RCX because of its wireless communication ability. However,
this limits the proximity sensors I can choose from. (Currently it uses
only two front bumpers.) For a long time I was thinking about switching
to Handy Board. (I don't mind using non-Lego electronic parts, but I'd
like to stick to Lego on the mechanical parts.)
This is why this laser idea is so attractive to me. I have already had
the hardware and the off-board image processing software in place, and
now I just need to arm the robot with two additional laser guns.
> You can use a 'bandpass optical filter' (get 'em at astronomy suppliers,
> get those that use c-mounts) to increase contrast without increasing
> processor load (remember, a 10mW laser is brighter than the sun at the
> suns surface PER AREA of emission).
Will this 'bandpass optical filter' interfere with the camera's current
duty as the barcode detector?
> Also remember that this sort of signal strength goes down by the 4th
> power, not the 2nd power per direct emission.
I understand that reflection doubles the distance the light travels, but
why does the signal strength go down by the 4th power instead of the 2nd?
> When I built this about 12 years ago for a project at a science museum I
> worked at we used a home built frame grabber and a C64 to do the numbers.
> The image processing will have to be done off-board of the RCX I suspect.
Now it is done off-board. I am toying with the idea of on-board
processing, as the image quality from my wireless camera (XCam2 through a
NTSC grabber) is quite poor. The problem is that most computers are too
heavy for the Lego motors to carry.
I am thinking about using the Handspring Visor with the Eye Module.
However, for barcode detection I estimate that it can only do one frame
for every one to two seconds. For laser interferometry alone it might be
able to get a more decent frame rate, but then for laser interferometry
alone this solution may be a bit overkill.
Thanks for your suggestions.
Cheers,
Hao-yang Wang
|
|
1 Message in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
