Hi there, is there any way to re-align the processed map with the base map?
It’s off by about 10m. This was my first flight with standard config using a Mavic Air.
The image dots are from the new map or the old? Do you have the capability to process with Ground Control Points? Not shoot them in the field just having the option in Upload…
The image dots are the images taken for the high res image you see on top. This is the first upload in the project so I guess the base map is just normal mapbox map. I did see the option for ground control points, wasn’t sure what that was for, I’ll have a play though. Is there anyway just to manually align it?
No way to manually align it in DroneDeploy. Just to be clear, the map that is shown is from those dots? There have been quite a few threads about the Mavic Air’s geotag coordinates being truncated to the last second which will cause shifts such as this. If you Google “mavic air geotag truncated” you may find a workaround.
With Ground Control Points the map would be fit to the coordinates of the base map. You can fake in GCPs with points created on visible objects from Google Earth or a GIS program like QGIS. This will only get you to about 3 meters accuracy though which is what you would get if the geotags weren’t truncated.
I believe there might be a firmware update out there that would fix this.
1 Like
Thanks heaps, I’ll have a look into the truncation, the seems pretty plausible. Yes, the map that is shown is from those dots. Really appreciate the help!
It looks as though the images were captured without the lat/long being truncated.
Was a good thought though!
2 Likes
Hmmm, the other thing that should have told me that wasn’t the root cause was that if they were truncated then there would have been no seperation in many of the points as 1 arc-second is about 100ft which means that all points in that second would be right on top of each other.
Please email support@dronedeploy.com. They need to look at the processing to see what happened. As I said before you will see shifts if this happens, but you will also see shifts up to 3m from map to map when everything is working correctly. This is a primary reason to use GCPs unless you are using the solution for Z-Axis management. I’d love to help you create some rough GCPs and reprocess similar to this thread, but we would use the horizontal data from one of the applications and the vertical from your map. I probably need to make a video of this as well.
If you want to try it out PM me a download of images. In DroneDeploy you can Export the Source Data then Share the map from the bottom left and I will be able to download them.
Turns out you can’t add GCP’s on the trial version. Interestingly I tried it again today (with a different project so it wouldn’t auto-align to the last one. And it’s still off but not as much and in a different direction, it’s really frustrating you can’t just manually align it!
I’ll leave it for now and try a few other mapping solutions, thanks heaps for your help though!
If you were to perform a manual upload using the same images, you are likely to get a wider processed map. It won’t of course stop the offset, but it would push it further out, allowing for the important boundary to be consistent.
This has worked well for me when I’ve needed more map, but have been restricted from flying beyond the boundary.
I also noticed the camera doesn’t take photos directly under the aircraft, it takes them at perhaps a 75 degree angle, is this normal? Drone deploy points the camera directly down as it takes off and gains altitude but as soon as it starts capturing, it’ll take the photos at an angle.
For example:
It does this in both the cross hatch mode and perimeter mode.
1 Like
Yes, crosshatch and perimeter are both oblique flight plans. When using crosshatch there are no nadir (bottom) images.
1 Like