Does anyone know how these LiDARs work? / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Does anyone know how these LiDARs work?

previous 1 2 3 4 next

Move Thread

LAN_403

Patrick

Fri Aug 12 2011, 10:37PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

BjÃ¸rn wrote ...

Yes, it is the phase difference that is recorded.

I have one of the Bosch units and it is very accurate but the refreshrate is about 3 Hz so it would be very slow for this use.

Yes this is the problem for our flying machines, the refresh rate (even at 40 Hertz) is just to slow. We also need them to weigh about 200 grams or less. The Dubai team also said they were signal noisy at they're devices quoted 10-12 hertz rate. (They only averaged in the real-world flight environment 6-8 hertz of useful signal. )

Carbon_Rod

Sat Aug 13 2011, 04:15AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

@Patrick
Hokuyo URGs:
* I seem to recall the scan rates can be set higher (see open driver for more details)
* A spinning prism & lenses sit on top of the sensor internally, but they are still unreliable outdoors
* Power consumption after spin-up is under a few hundred milliamps
* The surface being scanned can create "noise" (mirrors/glass etc. are fun), but the system noted above has some very nice filtering algorithms.
* IIRC they are not eye-safe at close proximity

Unlike several other manufacturers, Hokuyo will not support university student research initiatives.
You may have to get funding for your team with the help of your institution's faculty members.

Additionally, you may want to get someone to contact these people for you:

Cheers,

Patrick

Sat Aug 13 2011, 09:34AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Carbon_Rod wrote ...

@Patrick
Hokuyo URGs:
* I seem to recall the scan rates can be set higher (see open driver for more details)
* A spinning prism & lenses sit on top of the sensor internally, but they are still unreliable outdoors
* Power consumption after spin-up is under a few hundred milliamps
* The surface being scanned can create "noise" (mirrors/glass etc. are fun), but the system noted above has some very nice filtering algorithms.
* IIRC they are not eye-safe at close proximity

Unlike several other manufacturers, Hokuyo will not support university student research initiatives.
You may have to get funding for your team with the help of your institution's faculty members.

Additionally, you may want to get someone to contact these people for you:

Cheers,

we arnt outdoors, were in a arena with roof. about the noise, the staff of Iarc used foil tape and foil foam board to give us so difficulty.

We were all standing at close range, so thats kinda scary.

its good to know about my college staff/hokuyo.

Pinky's Brain

Sat Aug 13 2011, 04:23PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Just to be clear ... what you do is modulate the laser with say white noise (or a frequency swept signal) and then just correlate the received signal with the original to determine the phase difference of the modulating signal? (Rather than of the light itself, which becomes unmeasurable without some very high coherence length laser.)

Patrick

Sat Aug 13 2011, 07:11PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Pinky's Brain wrote ...

Just to be clear ... what you do is modulate the laser with say white noise (or a frequency swept signal) and then just correlate the received signal with the original to determine the phase difference of the modulating signal? (Rather than of the light itself, which becomes unmeasurable without some very high coherence length laser.)

Can anyone show a gragh or picture of how this is done? Im trying to wrap my mind around it, but this is hard to get. (Other websites seem to show "cartoonish" explanations of TOF only.)

Mattski

Sat Aug 13 2011, 07:54PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Patrick wrote ...

Can anyone show a gragh or picture of how this is done? Im trying to wrap my mind around it, but this is hard to get. (Other websites seem to show "cartoonish" explanations of TOF only.)

This is more or less what's going on in such a system. The original signal could be truly random or it could be a pseudo-random signal which has been chosen so that its self-correlation is very small when it is offset a small amount, and very large when it lines up exactly with itself.

Hope that makes sense.

Pinky's Brain

Sat Aug 13 2011, 08:19PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

They do it with a sweep, although personally I'd just use a repeated block of white noise and FFT based correlation followed by interpolation around the peak to get subsample accurate delay.

TOF usually refers to the use of really fast shutters to cut off pulses to determine distance by how much of a returned pulse arrives at a CCD before it cuts off by the shutter (it's compared against the received light from pulse not cut off by the shutter at all to compensate for surface reflectance). A very interesting method, but not really doable on a budget because of the shutter.

Patrick

Tue Nov 13 2012, 06:16AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

this site explains it well: ->

The sensor measures the phase shift between the transmitted and reflected signals. The picture shows how this technique can be used to measure distance. The wavelength of the modulating signal obeys the equation:

c = f âˆ™ Ï„

where c is the speed of light and f the modulating frequency and Ï„ the known modulating wavelength.

The total distance D' covered by the emitted light is:

D' = B + 2A = B + (Î¸ * Ï„) / 2Ï€

where A is the measured distance. B is the distance from the phase measurement unit. The required distance D, between the beam splitter and the target, is therefore given by

D = Ï„ * Î¸ / 4Ï€

where Î¸ is the electronically measured phase difference between the transmitted and reflected light beams.

It can be shown that the range is inversely proportional to the square of the received signal amplitude, directly affecting the sensorâ€™s accuracy.

[not sure what "n" is though?]

i wonder if a primative lidar could be made using a CD/DVD head... those have a laser, beam splitter and matched detector all in the same device, though the lens would probably need to be removed.

I presume that this is only good for 2pi or less of the beat frequency, or else it repeats on intervals.

here is the Neato XV-11 one ->

(it uses triangulation)

Pinky's Brain

Tue Nov 13 2012, 07:43PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

I don't see the need for the beam splitter ... easier to just measure the return pulse slightly off axis. Here is a nice page showing how probably most the range finders work (hell of a lot cheaper than using an ADC) :

Patrick

Tue Nov 13 2012, 11:37PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Pinky's Brain wrote ...

I don't see the need for the beam splitter ...

i think the splitter and emitter are meant to move so the rest doesnt have to.

ive been told that the XV-11 neato lidar (wich uses a pixel PSD, and line laser) use slip rings to conduct the signals from the rotating head to the stationary base.

Panavision 2048 pixel sensor...

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob