Hello: I’m currently processing data from a 3.2km long aqueduct. After breaking the project up into smaller sub-projects, I did initial processing of each sub-project. I then used manual tie points to join (merge) the subprojects together. The completed point cloud of the aqueduct is now “bowed” upwards at each end. This seems to be a textbook example of error accumulation when models are joined together in long strips. Can someone help me solve this issue?
I am guessing a bit here since I don’t have the Quality Report, pictures or any output but this sounds like a case where GCP vs. MTP makes all the difference. I think you could lock it down by switching them to GCP but it could be tricky without actual targets of survey points.
The “standard” response is that you need to use GCP ground targets. Another option is to process it all as one project…how many total gigapixels is the combined project because my Core i9 machine might be able to do it without sub-projects.