Replies to This Discussion

Permalink Reply by Jake Stew on July 18, 2012 at 4:33pm

No.  Nobody here uses lidar, 3DR doesn't make a lidar, and there are no commercial lidar solutions under $2k.

I've seen one post from someone using some sort of commercial lidar.  If you use the search feature you can probably find it.

All the lidar information on this forum will come up here...

http://diydrones.com/forum/topic/search?q=lidar

• ▶ Reply

Permalink Reply by Andrew Holdun on July 18, 2012 at 6:57pm

Why does no one use lidar here? Is it the expense or is it something else.Microdrones seems to have cracked a lot of the issues I am interested in dealing with. I did see a kick-starter for a cheap Lidar. Perhaps some people here would be interested in pursuing that?

http://www.microdrones.com/products/md4-1000/md4-1000-key-informati... 

• ▶ Reply

Permalink Reply by Cliff-E on July 18, 2012 at 7:41pm

http://diydrones.com/profiles/blogs/mits-quadcopter-with-laser

Too heavy, too expensive, needs a lot of CPU power. Maybe in a few years it will reach usability, especially with CPU power + SLAM.

• ▶ Reply

Permalink Reply by Jake Stew on July 18, 2012 at 10:24pm

Like Cliff said.

What are you actually trying to do?  In most cases there's no point to lidar and/or better solutions through sonar, barometric altitude, etc..

Someday I'll get into machine vision (stereoscopic 3D) if nobody beats me to it.  It's cheaper and way more powerful.  Lidar is a dead end I think.

• ▶ Reply

Permalink Reply by Andrew Holdun on July 18, 2012 at 11:10pm

I am trying to do this..actually something more commercial but think historic recreation of various buildings and sculptures and monuments. http://archive.cyark.org/

• ▶ Reply

Permalink Reply by Jake Stew on July 18, 2012 at 11:16pm

You want a machine vision solution then.  Lidar isn't going to work well enough for that.

• ▶ Reply

Permalink Reply by Andrew Holdun on July 18, 2012 at 11:26pm

Why wouldn't Lidar work for that and why is Machine Vision better. If so why isn't that used by say CyArk and in Visual FX to scan point clouds and recreate both geo and texture? Sorry if I am missing something. I am not too familiar with Machine Vision.

• ▶ Reply

Permalink Reply by Jake Stew on July 19, 2012 at 1:22pm

I've never seen a lidar system with the resolution and range you would need.  Even once you have the lidar data you still have to take photographs and map them onto the mesh.

Stereoscopic machine vision does all this at once.  Using two linked cameras you get a dual photo stream which the spatial data is computed from.

Lidar is going to be sensitive to movement and you'd need an ultra-high scan speed or need to be stationary.

The lidar units you're talking about are really expensive, heavy, need a lot of power, and need to be stationary. 

• ▶ Reply

Permalink Reply by Andrew Holdun on July 18, 2012 at 11:12pm

I am looking at some pretty heavy camera setups and have to think that the way the whole drone explosion has been going this will be do-able soon. I sure hope so.

• ▶ Reply

Permalink Reply by brakar on July 19, 2012 at 3:37pm

Andrew, it has been demonstrated that photogrammetry based on aerial photos from uav's & compact cameras can provide pointclouds of equal or better resolution/accuracy then traditional lidar.

Try google eg "apero micmac lidar"

• ▶ Reply

Permalink Reply by Andrew Holdun on July 27, 2012 at 10:13pm

How and who has demonstrated..any empirical evidence for lidar vs photogrammetry?

Any technical papers?

Thanks appreciate the comments and would love to read actual studies.

• ▶ Reply

Permalink Reply by Mike on July 28, 2012 at 10:21am

Mini-UAV-Borne LIDAR for Fine-Scale Mapping

http://tinyurl.com/d7opokw

• ▶ Reply

• ‹ Previous
• 1
• 2
• Next ›
• Page