The only thing PPK is going to do is insert more accurate geotags into the pictures. Is the topo for the developed areas, the tree areas or both? Are they expecting a DSM or DTM?
Hi Michael, thanks for always responding! You’re a great example of “sharing equals more”!
My first concern is whether this mission is feasible given that the drone is flying over dense trees and changing terrains. Is it safe? Need more specialized equipment?
The main goal is to identify any potential building sites and where roads would be the least expensive to put in (least steeps grades etc). So the client is wanting the ground topo in context of the boundaries and the exiting roads. They want to be able to use some key landmarks and get coordinates and map them to the DXF file and place that over a map. For example, overlay it on Google Map and county map.
DroneDeploy’s biggest limitation is going to be line of sight through the vegetation. Photographic systems cannot “see” through the foliage and therefore cannot provide elevation info for the ground surface. To get accurate ground elevation info I’m afraid LiDar or boots on the ground are the only solutions.
Or some point cloud editing… I’d be happy to help as a case study. Back to the site, can you share the coordinates or message them to me. I’d like to see the terrain and a couple of other map vantage points. I have a program that brings in the Google Earth terrain DTM.
Here is an example from the Carlson Precision 3D Topo software we use. Surface, draped tiles and 5/10ft contours. You can set the contours at whatever you want, but rendering time is highly dependent. Absolute accuracy is about 10m and relative accuracy is about 3m, but this is where I start with conceptual design. I can increase the resolution by about 300%, but for these purposes I don’t know that it would make much of a difference except for increasing the compile time. The dataset is from several sources from shuttle radar all the way down to Airborne LIDAR (3mA/1mR) if the area supports it. This helps us identify areas of interest that we can then go in a fly the drone to get better resolution. From What I have seen from this area now I think it could be done with photogrammetry, but it will be a task. I am guessing 75/75 standard and 65/65 crosshatch combined. The real work comes in at the point cloud editing and allot of trial and error on the bare earth filter. The last picture shows low-point detection which can be used to create 3D breaklines.
This is what I get when I bring it into CAD to start analysis and conceptual design. I can already start to see the flatter areas. In this case the bright green areas are flat and more reflective from the light source being straight above.
Last one. I would not fly this with DroneDeploy yet until terrain following is more dependable. Here is an example path that I would create in Google Earth at a constant 90m above ground that I would import into Litchi. Once again relying on 10m ground data so a little leeway is needed. Note that the elevation of the path is derived from the AGL at the waypoints so I would add intermediate points to follow the terrain even better.
DroneDeploy automatically handles multiple batteries although many of us like to force it to come home so it doesn’t re-shoot points. Once it passes a waypoint that you are pretty sure that it is not going to make the next leg (25% or so) force RTH.
For Litchi it’s not as elegant which is where some of the features are that make DroneDeploy beautiful. You just watch your battery percentage and RTH the same way. Then you can easily delete the waypoints you have crossed and restart the mission. At least in Litchi they are numbered.
Map Pilot will save you a lot of time. I don’t know the exact area of interest, but just grabbed a plot in the area and used terrain awareness and the mission created 209 waypoints for 60+ acres. I suppose if time is of no value, you could program Litchi to do it.
Hi Michael, I meant time savings in regard to placing all the waypoints in Litchi. Map Pilot uses SRTM data to program the waypoints relative to the surface.
Alternatively, You can fly a higher than normal mission and use the generated ortho to create your own terrain data which can be useful if the topography has changed significantly since the SRTM mission. I have not tried this yet myself.
Map Pilot is just another good tool to have in the box. It is ios only though.
Just checked it out and looks like a pretty cool program. Litchi uses the same data and you plan there or you can also plan in Google Earth. The one thing it doesn’t do is automatically space the pattern.
So long as you have a good connection to the controller, image capture spacing will probably be more accurate for overlap than a simple timing capture in Litchi too. I like Litchi too, btw !