Hello,
I tried merging two projects that were flown with overlapping coverage on the same day. Both projects have separate GCPs that were used to re-optimize the surfaces before they were merged. After merging the projects do not align and are shifted vertically from each other. This is odd to me as if I process the set of images from both flights together the produced surface is continuous…
Is it best to process imagery from different flights together even if they have two separate GCP setups?
Hi barrett,
Can you upload the quality report to this posting so we can take a look?
You have two viable options. You can process them separately and merge or process them as a single project. A single project is less work on your end, but if the dataset is quite large or requires different calibration settings for each subset, you might want to consider the merging technique.
Hi,
I had a similar problem, but in this case it was with a set of 20 GCPs, which only 2 appeared in both project: I added 3 MTPs to add more common features, but still it merged the point cloud to two very different heights.
After the umpteenth unsuccessfull merging process, I just used Rematch and Optimize on the merged .p4d, and I reached a good result.
Attached the final version, and the log, of the merged project:
231227_AleaPark_M.log (760.7 KB)
231227_AleaPark_M_report.pdf (1.5 MB)
And the log and report of the two starting projects:
231227_AleaPark_1 - copia.log (2.1 MB)
231227_AleaPark_1_report.pdf (1.9 MB)
231227_AleaPark_2.log (2.7 MB)
231227_AleaPark_2_report.pdf (3.2 MB)
As side note, I typically process the two input datasets in Pix4DCloud Advanced as a single flight, with excellent results, but this time I couldn’t: the first flight I forgot to remove the Dewarping option (in Pix4DMapper I can manually set the R-Parameters to 0, while in the Cloud it leads to a banana-shaped point cloud error), and the second flight I went in Manual Flight (the mission planner was too expensive in terms of battery duration)
Sincerely,
Giancarlo