Nice Dyl. Hey can you share the settings you use for the L2, and the path? I seem to miss a lot of data, area, detail, even using the L2. The automatic path generated by the DJI Pilot seems to have quite a bit of spacing between the runs and cuts the area a bit short, generating some circular scans as it turns at the ends of each run , and manually editing the path is something I haven't been able to do with the M350/L2 setup
Great tutorial Dylan !! Quick question. Are you releasing any LiDAR-related courses soon ? Seems like there's not enough training material out there for people wanting to get into the LiDAR field. I'm talking about thorough courses like detailed workflows, acquisition (settings, flying patterns etc...), processing, exporting and so on. Thanks man, keep up the great work !!! Merry Christmas !
Have a project that requires to scan a large 3 story building for the purposes of recreating Arch Plans “ as built”. No vegetation. Would photogrametry be enough? Thx
Yes it would work out really well! I have a dedicated workflow on my online course if you are interested that guides you on how to capture high-quality 3D models. Otherwise, capture lots of photos from as many angles as possible and you should get a good model.
In relation to building façade inspections, I assume that photogrammetry would be more effective for referencing specific areas of the façade? The example you provided, showcasing the exposed rebar on the silo, is an excellent illustration. However, I also need the capability to scan areas beneath tree canopies. I'm curious whether LiDAR can deliver image quality comparable to the one you demonstrated for the exposed rebar. Thanks for all the info in the video - I just subscribed to your channel.
Yes, but Ive just got L2, do you think I should get M3E for specific tasks or just try do it with l2 or instead of buying M3E get P1? Hmmm - considering for models / facades/ im surveyor proffesional licensed so you have a clue what Im doing mostly @@Louisianahelicam
@ I see you fixing with that big boy to that rock robotics are three v2 i’m saving up for that. It is the ultimate photogrammetry and LiDar combined game changer.
Can you elaborate on when you said the maps won't be "survey grade" since you aren't using GCPs? In talking with people in the industry about adding an RTK module to my drone, I've been told the RTK would reduce or completely eliminate the need for GCPs. If you had RTK during these maps, how accurate were they compared to a non-RTK flight? Thank you.
Any idiot could have seen that when Amazon shifted ALL their tech to video cameras and the software / recognition algorithms to use with them. Instead of planting rfid in their automated stores, they put cameras everywhere and progressed with making it so they could do just about EVERYTHING with just video :P
Wow ! Impressive differences ! So, photogrammetry is a better use if there's no vegetation ^^ And it's too bad that those technology cannot combine each other into a single camera. Maybe in few years... BTW thank you for the video ! I really enjoyed 👍👍
Incorrect. No need for Lidar in zero canopy area. Riegl Lidar systems start at $50,000 and unless you're using that you're not serious anyway. Besides, after exhaustive testing, I can get as good data from a $5000 setup as a $80,000 one. But most people won't test for years. They haven't got the time, money, patience or desire.
Seems to me it's not an either or situation, use sensor fusion! We were doing it on a small scale back in the 90s surveying days. Worked for a startup making laser survey hardware and software. Make a laser range map, take a bunch of pictures, map the pix onto the 3d range data. Took a whole day sometimes of computing time. Compute is so much cheaper today, and algorithms are better. Surprised you're not doing this, yet. ;-)
Good evening! I'm really interested in your videos and would love to help take your channel even further. My name is Val, and I'm a video editor who specializes in creating shorts and TikToks. I’d be happy to offer 3 free shorts as a sample of my work. If you like what I create, we can discuss potential collaboration moving forward. Looking forward to hearing from you!
Neither the answer for now is Gaussians. I would say get the data, use postshot to train it and go Gaussian splats. The difference is remarkable and life like environments.
Awesome work Dylan, best explaination/comparison by far.
Thanks!
Nice Dyl. Hey can you share the settings you use for the L2, and the path? I seem to miss a lot of data, area, detail, even using the L2. The automatic path generated by the DJI Pilot seems to have quite a bit of spacing between the runs and cuts the area a bit short, generating some circular scans as it turns at the ends of each run , and manually editing the path is something I haven't been able to do with the M350/L2 setup
I usually do a 35%-45% side overlap with 70% frontal. 8-15mph speed for capture and between 180ft-260ft agl
Thanks Dylan. Looks like I should pick up your course.
@FOX111975 it’s still on sale until Tuesday!
Great video brother, I need to get out there for some training soon. Great info and tutorial 👍
We’ve got more workshops coming up next year!
Excelent explanation and video! Congrats!!
Thank you!
Great tutorial Dylan !! Quick question. Are you releasing any LiDAR-related courses soon ? Seems like there's not enough training material out there for people wanting to get into the LiDAR field. I'm talking about thorough courses like detailed workflows, acquisition (settings, flying patterns etc...), processing, exporting and so on. Thanks man, keep up the great work !!! Merry Christmas !
Funny you bring this up, we are in full production of our LiDAR course that covers all of that and much more. Launching hopefully early Q1 next year!
@@DylanGorman awesome !! Looking forward !!
Have a project that requires to scan a large 3 story building for the purposes of recreating Arch Plans “ as built”. No vegetation. Would photogrametry be enough? Thx
Yes it would work out really well! I have a dedicated workflow on my online course if you are interested that guides you on how to capture high-quality 3D models. Otherwise, capture lots of photos from as many angles as possible and you should get a good model.
In relation to building façade inspections, I assume that photogrammetry would be more effective for referencing specific areas of the façade? The example you provided, showcasing the exposed rebar on the silo, is an excellent illustration. However, I also need the capability to scan areas beneath tree canopies. I'm curious whether LiDAR can deliver image quality comparable to the one you demonstrated for the exposed rebar. Thanks for all the info in the video - I just subscribed to your channel.
Yes
Yes, but Ive just got L2, do you think I should get M3E for specific tasks or just try do it with l2 or instead of buying M3E get P1? Hmmm - considering for models / facades/ im surveyor proffesional licensed so you have a clue what Im doing mostly @@Louisianahelicam
@@wzfellguard1 Not sure. I only use what I build for mapping
Very helpful video 👍🏼
This is a banger 👍😊
Thank you!
@ I see you fixing with that big boy to that rock robotics are three v2 i’m saving up for that. It is the ultimate photogrammetry and LiDar combined game changer.
How much can you charge for Lidar versus the photogrammetry?
Can you elaborate on when you said the maps won't be "survey grade" since you aren't using GCPs? In talking with people in the industry about adding an RTK module to my drone, I've been told the RTK would reduce or completely eliminate the need for GCPs. If you had RTK during these maps, how accurate were they compared to a non-RTK flight? Thank you.
Without GCPs, how would you know if RTK is accurate?
One or two checkpoints?
The title is good. When my friend bought a Faro scanner for $45k the rep said wait until photogrametry comes out. lol.
Oh boy lol, not what you want to hear after spending $45k
Any idiot could have seen that when Amazon shifted ALL their tech to video cameras and the software / recognition algorithms to use with them.
Instead of planting rfid in their automated stores, they put cameras everywhere and progressed with making it so they could do just about EVERYTHING with just video :P
Why use a point cloud, when you have access to a good GCP created 3D model, with no tree canopy's?
Wow ! Impressive differences ! So, photogrammetry is a better use if there's no vegetation ^^
And it's too bad that those technology cannot combine each other into a single camera. Maybe in few years...
BTW thank you for the video ! I really enjoyed 👍👍
You can 100% combine both technologies together!
The L2 has both LIDAR and an RGB camera. Thats how the point cloud is colourised.
Photogrammetry for everyday use. Lidar for professional use 👍
Incorrect. No need for Lidar in zero canopy area.
Riegl Lidar systems start at $50,000 and unless you're using that you're not serious anyway.
Besides, after exhaustive testing, I can get as good data from a $5000 setup as a $80,000 one. But most people won't test for years. They haven't got the time, money, patience or desire.
This is very true!
I appreciate this video!
Thank you!
good basic comparison , use case info
I am mapping/scaling/sizing whales over the ocean so which method is better?
Cheers much awesome tutorial well explained and I'm not even into mapping 😂🤔🤪❤
Do you see DJI coming out with a sub $2,000 Lidar drone for hobbyists?
AwsAWsome
No real answer, really depends on what your use case is and where/what you're mapping.
FOILAGE!! th-cam.com/video/Mqe8z0GmkXo/w-d-xo.html
lol
@@DylanGorman no offense man, love your videos and congrats on your business!
Seems to me it's not an either or situation, use sensor fusion! We were doing it on a small scale back in the 90s surveying days. Worked for a startup making laser survey hardware and software.
Make a laser range map, take a bunch of pictures, map the pix onto the 3d range data. Took a whole day sometimes of computing time. Compute is so much cheaper today, and algorithms are better. Surprised you're not doing this, yet. ;-)
I am using that exact scenario on a lot of my LiDAR projects. I just wanted to show the difference between the two techs separately
Hey cool, can anyone help me get started in this field. I'd owe you.......big time.....
I’ve got free and paid courses on my training platform if your looking for dedicated training
@DylanGorman yes I am thanks
Good evening!
I'm really interested in your videos and would love to help take your channel even further. My name is Val, and I'm a video editor who specializes in creating shorts and TikToks. I’d be happy to offer 3 free shorts as a sample of my work. If you like what I create, we can discuss potential collaboration moving forward.
Looking forward to hearing from you!
Until congress screes it
Neither the answer for now is Gaussians. I would say get the data, use postshot to train it and go Gaussian splats. The difference is remarkable and life like environments.
Not yet! As much as I love gaussian splats, they still have a long way to go before they are used for commercial grade work.
You can’t use the photogrammetry for anything and the lidar .. but I’m speaking from a visual look. I think the lidar is the way to go for data.
@@DylanGormannot to mention how for so many things you need a closed mesh. Not exactly something Gaussians are known for.
@lio1234234 exactly, gaussians are still cool though!