Hi Oto, good idea and really interesting. I did your exercise in a different way (even if i used your kml to identify the area): 1) in google earh i choose 4 point visible in the photo to use as GCP and so i created a gcp txt file 2) Import the txt file of GCP in CC 3) I launch the AT with rigid registration due to inaccurate GCP, 4) I changed the AT default setting ( keypoint density high, pair selection exhaustive, component construction mode multi pass), in this way one more photo was recognized in main component. and it works good... Here attached a file folder with GCP txt file some snapshot and 2 productions (3mx mesh file and kml mesh file). communities.bentley.com/.../example.zip
↧