Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

Determining altitude and spacing for surveys

Joined
Sep 7, 2017
Messages
50
Reaction score
6
Age
40
How does everyone determine altitude and spacing for surveys. I've done a couple now but it's been a bit of 'finger in the air' estimation when setting these parameters. Happy with the results but I'd be far more comfortable coming up with a way of setting them a bit more methodically.

For example, for a decent resolution and overlap for flying at 50m, what spacing would you use compared to that say at 30m for the same overlap?

The other thing I've been doing which seems to make a decent difference is checking the 're-fly at 90 degree offset' box but, when undertaking the flight, tilting the camera to 45 degrees to capture the side of buildings and structures. Does anyone else do this?... Is it the 'done' thing?
 
  • Like
Reactions: morry
I have not done any surveys yet so my reply will be from just from exploring this option. When you set up your survey it tells you at the bottom what will be ground resolution at the height you set in your survey. The number for the ground resolution can be seen as a text on the right hand side in your survey setting dialogue. It will say something like X cm per pixel. Since you know the resolution of your camera you can figure out what kind of space on a ground your pictures cover. As an example if the camera resolution is 5000x4000 pixel and you are getting a ground resolution of 1cm per pixel then your camera takes a snapshot of 5000x4000 cm or 50 x 40 meters space on the ground.

As for the spacing between survey flight paths I think DataPilot does it automatically based on your set altitude and an area on the ground that you want to survey. I believe that camera resolution settings are also taken into an account but it looks like it just takes it from your camera settings as there is no such a parameter in Survey settings.

And for the camera angle, it is just what we have been talking about for the last 1-2 months here. There are no such settings into Survey mode, It just flies 90 degrees angle by default.

I have H520 only for a couple of months and it has come with 1.3 version from a box but a lot of people say it was there before but disappeared from settings with 1.3 update. You can control a camera angle in waypoint mode only.

I hope Yuneec has not done removing a camera angle from survey mode intentionally and they will put it back. If you need it now try Pix4D capture app. It may be possible to do it through this app. It is free so it is worth a try. UGCS app is another option but it is not free,
 
In general, the lower altitude the better. Commercially, I understand 3cm/pix is normal, but the H520 is capable of much more. I am still using the 70% overlap as it seems to work well. Nadir is good for 2d surveys, but if 3D is wanted, a camera angle of 45 degrees is appropriate.
At present, DataPilot surveys are nadir, but the camera angle can be changed by inserting a waypoint before the survey. The camera will then retain this angle throughout the survey. They could have made this easier.
As Wingman correctly points out, obliques were previously part of the DataPilot firmware, but now we have a workaround.
The E90 camera is excellent for surveys since the option to remove distortion appeared in the camera menu and I have not found that reflying nadirs at 90 degrees offers any significant improvement.
Pix4D is claimed to be capable of handing photos taken at any angle, but I have only used 90 and 45 degrees. Other firmware may differ.
 
  • Like
Reactions: emilereynolds
Thanks for the comments guys. Haven't had chance to look into them properly yet but definitely will - just wanted to pop in to say thanks... I'll let you know how I get on!
 
In general, the lower altitude the better. Commercially, I understand 3cm/pix is normal, but the H520 is capable of much more. I am still using the 70% overlap as it seems to work well. Nadir is good for 2d surveys, but if 3D is wanted, a camera angle of 45 degrees is appropriate.
At present, DataPilot surveys are nadir, but the camera angle can be changed by inserting a waypoint before the survey. The camera will then retain this angle throughout the survey. They could have made this easier.
As Wingman correctly points out, obliques were previously part of the DataPilot firmware, but now we have a workaround.
The E90 camera is excellent for surveys since the option to remove distortion appeared in the camera menu and I have not found that reflying nadirs at 90 degrees offers any significant improvement.
Pix4D is claimed to be capable of handing photos taken at any angle, but I have only used 90 and 45 degrees. Other firmware may differ.
Hi, is the UGCS app fully supported by Yuneec 520 and is it superior to DataPilot?
 
In general, the lower altitude the better. Commercially, I understand 3cm/pix is normal, but the H520 is capable of much more. I am still using the 70% overlap as it seems to work well. Nadir is good for 2d surveys, but if 3D is wanted, a camera angle of 45 degrees is appropriate.
At present, DataPilot surveys are nadir, but the camera angle can be changed by inserting a waypoint before the survey. The camera will then retain this angle throughout the survey. They could have made this easier.
As Wingman correctly points out, obliques were previously part of the DataPilot firmware, but now we have a workaround.
The E90 camera is excellent for surveys since the option to remove distortion appeared in the camera menu and I have not found that reflying nadirs at 90 degrees offers any significant improvement.
Pix4D is claimed to be capable of handing photos taken at any angle, but I have only used 90 and 45 degrees. Other firmware may differ.
Is that correct? If you insert a waypoint with a camera angle of 45 degrees before the survey grid, the 45 degree angle will be retained. I understood that any survey grid would default to nadir? This would be great and a simple workaround.
 
Is that correct? If you insert a waypoint with a camera angle of 45 degrees before the survey grid, the 45 degree angle will be retained. I understood that any survey grid would default to nadir? This would be great and a simple workaround.

It is indeed correct and unfortunately because you could easily determine in the survey setup the camera tilt previously :cool:
 
  • Like
Reactions: Edb and Graham
It has worked for me several times. I prefer DataPilot to the PIX4D app which I find difficult to read due to the colours of the text and small buttons that don't suit my large fingers. Keep the waypoint dodge quiet or Yuneec might remove it at the next update.;)
 
  • Like
Reactions: Edb
In general, the lower altitude the better. Commercially, I understand 3cm/pix is normal, but the H520 is capable of much more. I am still using the 70% overlap as it seems to work well. Nadir is good for 2d surveys, but if 3D is wanted, a camera angle of 45 degrees is appropriate.
At present, DataPilot surveys are nadir, but the camera angle can be changed by inserting a waypoint before the survey. The camera will then retain this angle throughout the survey. They could have made this easier.
As Wingman correctly points out, obliques were previously part of the DataPilot firmware, but now we have a workaround.
The E90 camera is excellent for surveys since the option to remove distortion appeared in the camera menu and I have not found that reflying nadirs at 90 degrees offers any significant improvement.
Pix4D is claimed to be capable of handing photos taken at any angle, but I have only used 90 and 45 degrees. Other firmware may differ.

I was an early adopter of the 520 one of the first to get it in the UK.
I still have the 1.19 datapilot app for the iMac (unfortunately not for PC as I don’t have one) and use that to program surveys and change the camera angles instead of the waypoint method. I then save this as a plan and upload as normal to st16s
If anyone wants this version let me know.
The angles settings disappeared from 1.2.5 onwards
 
In general, the lower altitude the better. Commercially, I understand 3cm/pix is normal, but the H520 is capable of much more. I am still using the 70% overlap as it seems to work well. Nadir is good for 2d surveys, but if 3D is wanted, a camera angle of 45 degrees is appropriate.
At present, DataPilot surveys are nadir, but the camera angle can be changed by inserting a waypoint before the survey. The camera will then retain this angle throughout the survey. They could have made this easier.
As Wingman correctly points out, obliques were previously part of the DataPilot firmware, but now we have a workaround.
The E90 camera is excellent for surveys since the option to remove distortion appeared in the camera menu and I have not found that reflying nadirs at 90 degrees offers any significant improvement.
Pix4D is claimed to be capable of handing photos taken at any angle, but I have only used 90 and 45 degrees. Other firmware may differ.
Hi Graham, On Monday 18th I'm heading to Yuneec in Romford for a H520 demo. I do have some scepticism & questions for them, I did have a phone conversation with them today and they are adamant that new upgrades are on the way for the H520 within months!! Hmmm anyway, as I am very new to drones can you give me a bit of a steer on the kind of information I need to draw out of them and the kind of questions, using your experience, I need to be asking? you help on this is greatly appreciated my email is [email protected]
 
How does everyone determine altitude and spacing for surveys. I've done a couple now but it's been a bit of 'finger in the air' estimation when setting these parameters. Happy with the results but I'd be far more comfortable coming up with a way of setting them a bit more methodically.

For example, for a decent resolution and overlap for flying at 50m, what spacing would you use compared to that say at 30m for the same overlap?

The other thing I've been doing which seems to make a decent difference is checking the 're-fly at 90 degree offset' box but, when undertaking the flight, tilting the camera to 45 degrees to capture the side of buildings and structures. Does anyone else do this?... Is it the 'done' thing?

I’ll make it simple (oblique sample only)

1. Fly two patterns and alt.

2. Determine your waypoints and set your alt and angles accordingly.
(Give or take. For example my structure is 80ft without obstructions) where’s my plan.
I will map it out as follows:
a. Approximately 150ft Nadir 0 deg
b. Approximately 80ft Oblique 45 deg
c. Approximately 50ft or 30ft Oblique 80 deg
d. Groupings altitude and camera angle varies you’re now taking taking photos of textures and details.

* Imagine a lazy Susan turntable: Fly in orbit mode or take a photo manually in a circular grid pattern via smart or intelligent mode.

The basic of mapping is basically you’re taking photos in all sides and hit all the points. You want also achieve a 70% overlap lesser than this, you increase your altitude, but never go below 60% on your side and front overlap ratio or else, you will be spending more time in mesh lab filling in holes.

3. The other method is try to imagine a box and put your model in the center of that box and take a photo of it in all different points.

4. Take it home: Adobe Recap Pro is by far the best, and it’s the same programming software used by Drone Deploy that is why I stop using DD and other apps in the market.

Learn the basic of mapping, take a photo of a fruit using your camera phone or dslr and load it using in your 3d mapping software (Agisoft or DD) so when you fly on a job-site you know exactly what to do and expect.

Best of luck.
 
My recent map, flown at 260ft and 150ft. I had no choice to go lower, due to high tension wires, and freeway overpass.

IMG_8129.jpgIMG_8130.jpg

The model is not pretty and I will be retaking this assignment using a P4P.
 
My recent map, flown at 260ft and 150ft. I had no choice to go lower, due to high tension wires, and freeway overpass.

The model is not pretty and I will be retaking this assignment using a P4P.

Good points to consider, but curious why your sample turned out the way it did? Here is a development of about a 60 acre site. Flown nadir in 1 rectangle pattern at 250' using 1 battery and consisting of just 221 images. Final ground resolution of 1.94cm/pix. 520 with e90 camera:
 

Attachments

  • OrthoSM.jpg
    OrthoSM.jpg
    3.3 MB · Views: 37
  • Close up.JPG
    Close up.JPG
    253 KB · Views: 35
  • 3D.JPG
    3D.JPG
    911.4 KB · Views: 38
Good points to consider, but curious why your sample turned out the way it did? Here is a development of about a 60 acre site. Flown nadir in 1 rectangle pattern at 250' using 1 battery and consisting of just 221 images. Final ground resolution of 1.94cm/pix. 520 with e90 camera:

I see what you are doing and the alt that you are in. How accurate is your GCPS? Are you mapping in relative or absolute. Our subject is in relation to an oblique structure. If it’s the sharpness of the photo, it’s an iPhone screen shot. I will post a better one. Let me see a zoomed perspective at 45 deg of your material,’specially on the structures.
 
I see what you are doing and the alt that you are in. How accurate is your GCPS? Are you mapping in relative or absolute. Our subject is in relation to an oblique structure. If it’s the sharpness of the photo, it’s an iPhone screen shot. I will post a better one. Let me see a zoomed perspective at 45 deg of your material,’specially on the structures.
I wasn't degrading what you were doing, just curious why the sample you showed lacked detail, given that they were flown at about the same altitude. The sample I posted was not flown for oblique or 3D effect, yet I thought was impressive for a single pass nadir set that shows what can be achieved from the e90. This was simple survey mission from the DataPilot program. All defaults kept and at 70% overlap. Report attached..
 

Attachments

  • Report.pdf
    14.6 MB · Views: 22
I wasn't degrading what you were doing, just curious why the sample you showed lacked detail, given that they were flown at about the same altitude. The sample I posted was not flown for oblique or 3D effect, yet I thought was impressive for a single pass nadir set that shows what can be achieved from the e90. This was simple survey mission from the DataPilot program. All defaults kept and at 70% overlap. Report attached..

I understand that you weren’t. One pass at that alt 250 photos 100+ points? Your map is impressive, I specially like the stock pile.
 
I understand that you weren’t. One pass at that alt 250 photos 100+ points? Your map is impressive, I specially like the stock pile.
Yes, at the time that is what impressed me as well. To zoom in that far and see the accuracy of the stitching, I think speaks most to the quality of the camera.
 
  • Like
Reactions: record.play.rewind
These are from the H+ flown in manual grid at at 200ish ft overlap varied some are from 70%-80% one battery
 

Attachments

  • 2018-09-13 LENNAR-1 YUNHPLUS.jpg
    2018-09-13 LENNAR-1 YUNHPLUS.jpg
    678.2 KB · Views: 38
  • 2018-09-13 LENNAR-2 YUNHPLUS.jpg
    2018-09-13 LENNAR-2 YUNHPLUS.jpg
    770.4 KB · Views: 38
  • 2018-09-13 LENNAR-3 YUNHPLUS.jpg
    2018-09-13 LENNAR-3 YUNHPLUS.jpg
    697.4 KB · Views: 34
  • Like
Reactions: 10-8
And see the screenshot photos that I took straight from my phone and posted here.. Those are ugly! LOL
 
  • Like
Reactions: 10-8

New Posts

Members online

Forum statistics

Threads
20,955
Messages
241,599
Members
27,284
Latest member
csandoval