I've been tasked with doing something that seems impossible, so here goes:
Goal: to have a drone scan an area based on a pre-determined pattern, "automatically" assemble those scans, and (eventually) computationally compare the two scans for differences. FWIW: the test area will be a reasonably tightly packed factory (lots of cables randomly hanging down, tightly packed equipment, walls that are not necessarily consistent), and the changes will mostly be fork trucks, humans, boxes ... not the stuff I actually want (overhead steel, equipment).
The first step is finding a drone that can do the scanning part, with reasonable accuracy, and as automatically as possible (even if that is accomplished at the software end). Accuracy is not of paramount importance, but reliability, and providing an assembled final scan? That is.
Does there exist a drone + LIDAR that can accomplish this? Am I overcomplicating things with the LIDAR requirement? Any insight would be appreciated. I am reasonably new to the LIDAR game (BLK360).
"Daily" Laser Scanning
-
- I have made 10-20 posts
- Posts: 18
- Joined: Tue Sep 15, 2020 12:07 am
- 3
- Full Name: E Casagrande
- Company Details: Tandem
- Company Position Title: Tandem
- Country: USA
- Linkedin Profile: No
- Has thanked: 17 times
- Been thanked: 2 times
-
- V.I.P Member
- Posts: 544
- Joined: Fri Feb 24, 2017 10:47 am
- 7
- Full Name: Martin Graner
- Company Details: PointCab GmbH
- Company Position Title: Research and Development
- Country: Germany
- Linkedin Profile: No
- Has thanked: 160 times
- Been thanked: 175 times
- Contact:
Re: "Daily" Laser Scanning
I'd try to reduce the dimension (given this is possible).
Use a drone with a camera rather than a laser scanner, then use Photogrammetry software to calculate the images together to a true orthophoto.
Comparing two images for changes of larger areas is more easy than comparing point clouds for changes (especially since it sounds like your scans will be very noisy).
Bonus: you can calculate a point cloud / mesh out of the images.
Use a drone with a camera rather than a laser scanner, then use Photogrammetry software to calculate the images together to a true orthophoto.
Comparing two images for changes of larger areas is more easy than comparing point clouds for changes (especially since it sounds like your scans will be very noisy).
Bonus: you can calculate a point cloud / mesh out of the images.
-
- V.I.P Member
- Posts: 205
- Joined: Mon Feb 25, 2019 4:42 am
- 5
- Full Name: Major Domo
- Company Details: VeritasVfx
- Company Position Title: ceo
- Country: Portugal
- Linkedin Profile: No
- Has thanked: 17 times
- Been thanked: 53 times
Re: "Daily" Laser Scanning
Why a drone?ecasagrande wrote: ↑Fri Jan 15, 2021 4:07 pm I've been tasked with doing something that seems impossible, so here goes:
Goal: to have a drone scan an area based on a pre-determined pattern, "automatically" assemble those scans, and (eventually) computationally compare the two scans for differences. FWIW: the test area will be a reasonably tightly packed factory (lots of cables randomly hanging down, tightly packed equipment, walls that are not necessarily consistent), and the changes will mostly be fork trucks, humans, boxes ... not the stuff I actually want (overhead steel, equipment).
The first step is finding a drone that can do the scanning part, with reasonable accuracy, and as automatically as possible (even if that is accomplished at the software end). Accuracy is not of paramount importance, but reliability, and providing an assembled final scan? That is.
Does there exist a drone + LIDAR that can accomplish this? Am I overcomplicating things with the LIDAR requirement? Any insight would be appreciated. I am reasonably new to the LIDAR game (BLK360).
Seems like a handheld scanner is what you should be thinking.
Establish a control network and you should be sorted in double time.
If a drone must be used, then VXGrid pointed you to the right solution, skip the lidar and go with photogrammetry if the lighting allows it. Same story about control network.
Allow for a beefy processing machine. Creating a point cloud from photos can be super painful if you cheap out on it.
-
- V.I.P Member
- Posts: 1237
- Joined: Mon Jan 04, 2010 7:51 pm
- 14
- Full Name: Jed Frechette
- Company Details: Lidar Guys
- Company Position Title: CEO and Lidar Supervisor
- Country: USA
- Linkedin Profile: Yes
- Location: Albuquerque, NM
- Has thanked: 62 times
- Been thanked: 220 times
- Contact:
Re: "Daily" Laser Scanning
I'll suggest a third route: Turn the entire factory in to one big scanner by permanently mounting a bunch of low cost lidar sensors, such as the ones from Ouster, throughout the environment. The initial engineering investment will be higher to begin with but with everything hardwired and permanently in place it will be much easier to do automatic scanning and analysis.
A few years ago we worked up a proposal to do something similar for a smaller environment with repeat scans every minute and similar things are done at a much larger scale in some open pit mines so it is definitely doable.
A few years ago we worked up a proposal to do something similar for a smaller environment with repeat scans every minute and similar things are done at a much larger scale in some open pit mines so it is definitely doable.
Jed
- smacl
- Global Moderator
- Posts: 1409
- Joined: Tue Jan 25, 2011 5:12 pm
- 13
- Full Name: Shane MacLaughlin
- Company Details: Atlas Computers Ltd
- Company Position Title: Managing Director
- Country: Ireland
- Linkedin Profile: Yes
- Location: Ireland
- Has thanked: 627 times
- Been thanked: 657 times
- Contact:
Re: "Daily" Laser Scanning
Not sure how well this would work for the hanging cables. I've been involved in a lot of drone work in the railway industry recently and the photogrammetry results for overhead lines has been terrible. I'd imagine you'd need multiple passes with different camera angles to get a decent result which could add significantly to capture and processing costs. Something like a NavVis VLX might be a better bet if it has adequate visibility of all the objects you're trying measure.
-
- V.I.P Member
- Posts: 544
- Joined: Fri Feb 24, 2017 10:47 am
- 7
- Full Name: Martin Graner
- Company Details: PointCab GmbH
- Company Position Title: Research and Development
- Country: Germany
- Linkedin Profile: No
- Has thanked: 160 times
- Been thanked: 175 times
- Contact:
Re: "Daily" Laser Scanning
I definitely lack the practical experience here, so thanks Shane for pointing that one out.smacl wrote: ↑Sun Jan 17, 2021 11:29 amNot sure how well this would work for the hanging cables. I've been involved in a lot of drone work in the railway industry recently and the photogrammetry results for overhead lines has been terrible. I'd imagine you'd need multiple passes with different camera angles to get a decent result which could add significantly to capture and processing costs. Something like a NavVis VLX might be a better bet if it has adequate visibility of all the objects you're trying measure.
The other question is how well everything will get assembled in that environment with all the moving people.
Perhaps the idea from Jed would be the best go?
- smacl
- Global Moderator
- Posts: 1409
- Joined: Tue Jan 25, 2011 5:12 pm
- 13
- Full Name: Shane MacLaughlin
- Company Details: Atlas Computers Ltd
- Company Position Title: Managing Director
- Country: Ireland
- Linkedin Profile: Yes
- Location: Ireland
- Has thanked: 627 times
- Been thanked: 657 times
- Contact:
Re: "Daily" Laser Scanning
Jed's solution certainly sounds interesting and isn't one I would have thought of. For drone based I'd be looking at Stormbee or drone mounted ZEB horizon. Having looked at data from both of these solutions it is good but not as accurate as terrestrial scan data. @E Casagrande, you'd definitely need to get the vendors in to do a proof of concept here to make sure the results met your requirements in terms of accuracy and scan density.