It seems as if something in the DD mapping engines has changed in the last 3-4 months that is resulting in skewed orthomaps. By skewed, I mean when I bring the tif into civil 3D an compare to site plans, one side of the site may be aligned within a foot but the other side of the site will be off by 7-8 feet. When we use GCPs with the same flight characteristics, the issue is resolved. The problem is we are not enterprise and paying $50 for GCPs every time we process a map (weekly) is not realistic.
I can’t for the life of me understand why DD insists on charging for GCPs, it does nothing but force customers to consider competitors when if they didn’t they would be the clear choice, but i digress.
This particular site is 60 acres, relatively flat, and has ongoing construction. I have tried everything that I can think of to get a better end product, hoping someone more knowledgeable might have suggestions.
I’m flying a Mavic 2 Pro, 80/80 overlap.
I’ve tried crosshatch w/perimeter, crosshatch w/o perimeter, no crosshatch w/perimeter, no crosshatch no perimeter. There are definitely noticeable differences in 3D but I’m really only concerned with the orthomap for this project and the various configurations offered little/no changes in results other than size and fly/upload/processing time. Is anyone else seeing this issue? Any other ideas on things to try?
Thanks!
1 Like
A 2-3m shift is pretty typical without PPK, RTK or GCP’s.
While I do agree that DroneDeploy’s pricing structure could be better and GCP’s should definitely be included in the Business plan it’s not really an apples-to-apples comparison between DroneDeploy and other solutions when you look into the details. Maybe another solution would fit your use-case better.
Agree that a shift of that magnitude is typical and can definitely work with that. The skew is what is causing problems.
Are you sure it’s not a shift and rotation? Or are string dimensions from the “good” side to the “bad” side not matching? If you have a successful map do some two-tape dimensions from two good points to one bad point. Can you share a good and bad map?
Hey Michael, sorry for the slow response. I’m finally having time to get back to this issue. Here is what I think is a pretty good example of the skew I’m talking about.
Note this is a non-GCP map. The pink points were recorded with a total station. Using the well and the front stoop to align the map. I put a pk nail in the stump in the top left of the image labeled CTRL4 and another nail in the stump by the road labeled CTRL3. You can see that although some of the map appears to be aligned, the 101-105 edge of road points and CTRL3 point are several feet off. I’ll do another flight to capture the painted controls but thought this represented the issue pretty well.
Is this loaded into CAD or aligned on DroneDeploy?
imported as a geotiff in Civil 3D
1 Like
I would expect the rotation and placement to be off without GCP’s, but not distorted. Is this a 0,0 import or did you move it to match CTRL4?
Even if you stretched the drawing out to match CTRL3 to the stump point 102 would be off the pavement. Can you try a 3-point alignment of the image to the points? I am curious to see how the overall map scales.
