Terrain capture sand area with rocks

Hello, I have to capture with high detail a large area that contains rocks and saνd is near the sea.

What is the best method?

Combining orbit flights and grid flights with camera to be in 60 degrees?

For now I have a phantom 4 pro?

 May I need RTK version of phantom for best results?

Ι was thinking also to add a terrestrial laser scanner. 

I know that context capture has some problems aligning images when mixing orbit and grid flights sometimes.

Parents
  • How big is a 'large area'? What kind of resolution do you need? Are there lots of rocks, or is it mostly sand? If there are enough rocks to have some in each photo you'll be fine. I would start with a nadir double grid 'up high'. That might be 60 meters or so. Then I would come down 25% of the way to the ground and do oblique double grid with the gimbal at 70 degrees. Maybe do another run 25%  closer. Keep doing that until you get the ground sampling distance that you're looking for. 

    You definitely don't need RTK.

    ContexCapture can give you more resolution than a laser scanner if you have lots of pixels and you get close. 

    Context capture doesn't struggle with mixing orbits and grids, but it does struggle with changing your distance to the subject by more than 75% or so. 

  • The area is 35.000 square meters the resolution that we need is 1cm per pixel or less.

    The rocks is along the sand.

    Ι was thinking to begin from 50m and go down to 10 meters in the double grid and nadir with many free flighs lower to cover areas near the ground or under the rocks. 

    Maybe a Mavic 3 pro will give bteer results?

    Also thinking to add a laser scanner in order to give more accuracy to the 3d model and cover some holes.

    Any other thoughts will be appreciated.

  • That isn't that big of an area in the grand scheme and the resolution isn't super high either. The flight app that you are using should tell you what your resolution is at different altitudes. The DJI app for the Phantom does, I believe. I'm not sure what altitude will give you 1 cm with the Phantom, but it's pretty high. Probably over 60 m. Maybe even 100 m. You might even be able to do the project in one flight and still get the 1 cm resolution. There is no reason to get much lower unless you are trying to capture details of the sides of the rocks. I have had good luck with a 70 degree gimbal angle. It catches the sides of things and also gets some nadir view in the field of view. CC does well with it. If you need more detail of the rocks I would suggest flying a 25% lower double grid flight just over them. 

    It sounds like there is sand along the water with an area of rocks on the land side. The sand by itself is going to be pretty featureless and that will make it hard to model. Flying high enough to get some rocks into each photo will help CC to tie everything together. A bunch of photos of just sand probably won't have enough features to work. Flying at 30 m will probably give you lots of useless images of just sand and CC will probably kick them out. If it doesn't matter to the project, you could draw some features in the sand and that would help CC to find matching pixels. 

    The Phantom has a 1" CMOS sensor. The Mavic 3 has a 4/5 CMOS. Both are 20MP. I would argue that the Phantom has a slightly better camera system although I bet you couldn't tell the difference. There's no reason to buy a new drone in this case. 

    Laser scanning this project would be a challenge. To get a big, featureless, outdoor area to register you would probably have to use targets (unless there were buildings close to the rocks). Best practice would be to locate them with a total station. You would want a scanner with some real range. You would also need ground control that was pretty accurate to get the scan point cloud to register to the mesh. Then there's the learning curve for doing that. 

    This is a pretty straightforward project. Get a double grid up high, capture the detail in the rocks if you need it and stuff it into CC. You'll be fine. 

    Just out of curiosity, what is the purpose of the project?

  • Terrestrial scanner is not suited for capturing terrain either as the angles are small and can't capture with good accuracy and density. This is definitely very well adapted for photogrammetry and there will be no holes.

    You could try this app(free trial) which helps to visualize the mission plan - https://app.droneharmony.com/

    You could fly double grid around 50-70m high. No need to go lower. Also may not need to fly all area in double grid would be enough to some in double grid and some using just grid with angled camera. If you need better resolution for some rocks or then go lower to them by spiraling down and shooting this way it won't cause issues with splitting in multiple components.

Reply
  • Terrestrial scanner is not suited for capturing terrain either as the angles are small and can't capture with good accuracy and density. This is definitely very well adapted for photogrammetry and there will be no holes.

    You could try this app(free trial) which helps to visualize the mission plan - https://app.droneharmony.com/

    You could fly double grid around 50-70m high. No need to go lower. Also may not need to fly all area in double grid would be enough to some in double grid and some using just grid with angled camera. If you need better resolution for some rocks or then go lower to them by spiraling down and shooting this way it won't cause issues with splitting in multiple components.

Children