DIY LiDAR Thread

[Addison Snyder]

Starting this dedicated thread to discuss underwater LiDAR and making custom/DIY tools for mapping underwater stuff.

Seems to me that most pre-built LiDAR systems run in the range of $10k+, especially if built for underwater purposes. I'm attempting to make a fairly simple (by comparison) system using a laser rangefinder, an encoding motor, and a slip ring to allow for connections onto the spinning motor shaft and attached rangefinder. The laser rangefinding module I've ordered uses a 635nm (red visible) laser, rather than the common IR used in most consumer rangefinders. Ideally, the laser would be blue or some wavelength less absorbed by water, but I can't find any. So I'm going to stick to the red laser for a POC above water, probably test how far (if at all) it works underwater, and go from there. I'm still figuring it all out, but I'm going to try to run this whole thing off of a raspberry pi and arduino, later adding depth sensors and other stuff to estimate positioning (lol anyone have a spare laser gyrometer?).

My end goal is to log data-points to a file where eventually it will be turned into a crude 3D cross-section of a cave, open-sourcing it and the data under a GPL-style license. People don't need a pixel-perfect map of a cave section to benefit, so precision isn't a total priority.

Aside from opening a dedicated LiDAR thread and spitting out lofty aspirations, anyone wanna colab on this? Thoughts? I've got a million ideas swimming around for how to guesstimate position/velocity, but the cross-section LiDAR comes first.

 

[Johnoly]

.. so precision isn't a total priority.

Divers are still looking for 100% refraction error correction in a silted out cave that can be mounted on an small ROV. If precision is not a priority, there are dozens of cheaper methods to map a cave than lidar. I hope you find a cost effective precise method.

 

[-JD-]

Sounds interesting.

Some searching revealed some 520nm Green modules from JRT in China:
Green Laser Distance Module Manufacture and Green Laser Distance Module Supplier in China
Oem 60m Green Laser Distance Meter Module - Buy Green Laser Distance Meter,Laser Module Green,Oem Laser Distance Module Product on Alibaba.com

 

[lostsheep]

Subd

 

[Addison Snyder]

Divers are still looking for 100% refraction error correction in a silted out cave that can be mounted on an small ROV. If precision is not a priority, there are dozens of cheaper methods to map a cave than lidar. I hope you find a cost effective precise method.

Not sure I understand the refraction part. Are you talking about light going from the inside of an air-filled tube to water? I plan for the rotating laser to pass through a clear acrylic tube at a ~90° angle, and am pretty naively just hoping it works out. If I can get readings +/- 6 inches from a yard away I'll be happy. I'm guessing it'll be more of a it-works-or-it-doesn't sort of thing though. Probably will get complicated. What cheaper methods are you referring to? I've explored sonar as well, particularly the Ping 360. The thing that disqualifies that (in my eyes) is that A) minimum reading is .75m and B) it takes too long to do a scan (9 seconds for a 360° sweep with any reasonable precision). It's also $2500. I can get an industrial-grade equivalent with better minimum reading for somewhere around $10k, but the sweep times are still slow. I'm trying to keep the budget in the low 4-figures.

 

[Addison Snyder]

Some searching revealed some 520nm Green modules from JRT in China:
Green Laser Distance Module Manufacture and Green Laser Distance Module Supplier in China
Oem 60m Green Laser Distance Meter Module - Buy Green Laser Distance Meter,Laser Module Green,Oem Laser Distance Module Product on Alibaba.com

Sick! Going to go for one of these if I hit wavelength-related limitations on what I already ordered. Although I need something I can poll more than 3 times a second. I've seen a lot of laser rangefinders that are limited to 3hz, and then some that advertise 100hz.

 

[The Ruttmeister]

Any reason why lidar and not structured light?

There are lots of cheap modules using structured light for 3D scanning. It would probably be easier and cheaper to use several of those to cover 360 degrees, than going full lidar. Unless you need the long range?

 

[Addison Snyder]

Any reason why lidar and not structured light

Initial reason for going with LiDAR was the simplicity of the components and data. I'm eventually hoping to mount something on the front of a DPV, figuratively hit record, travel through a cave, and take the combined data points and turn them into a (crude) 3D map. Structured light seems (at my first glance) very software intensive and more tuned to scanning localized objects. Maybe I'm wrong though. Here's a cool paper.

There are lots of cheap modules using structured light for 3D scanning. It would probably be easier and cheaper to use several of those to cover 360 degrees, than going full lidar. Unless you need the long range?

Do they work off of the visible spectrum? Any links or examples you had in mind?

 

[wetb4igetinthewater]

@Addison Snyder

Have you spoken to Natalie Gibb from Under the Jungle in Mexico? If I'm not mistaken, she's doing a bit of mapping of caves in the Yucatan.

I have shared a link to this thread with Norsk Remote Sensing (Norsk Remote Sensing) who is doing the laser scanning on my upcoming project (HAES | Hellenic Archaeological Exploration Society). I'm curious with what he has to say. When we had dinner last Friday, he mentioned that the critical part is the data processing, which is why the results of photogammetry are so poor when looking close up. That's one of the services he provides as a consultant. Of course, for purposes like mapping a cave, that isn't a big deal, but it is more of an issue of a shipwreck.

EDIT: photogammetry with good post processing is probably your best bet.

 

[Rick Brant]

I spent some time looking at underwater LIDAR for use in scanning reef areas but found that there were no great options. I'm using photogrammetry now instead and it is working very well. Of course you would need light to work in a cave, but we also sometimes use lighting for reef areas.

Most reef photogrammetry now is done using Agisoft's MetaShape, but Apple released some very nice system-level support for photgrammetry a year or so ago. It's now very easy to construct 3D models from images captured by an iPhone in a dive housing. I have released an iOS app -- ReefScan -- for capture and a MacOS app -- ReefBuild -- for building the models. They are very easy to use and I am working on several features to make them even more useful for reef surveys. A similar application-specific model could be used for caves.