Good Day to All,
In an aerial survey, I have taken a smaller resolution set of photos of an entire area, A, to be surveyed. Right after that, I have flown (with the same camera, lens, etc.) over a smaller area B (comprised in area A) to take higher resolution photos from a lower height AGL. Surveyed ground control points (GCPs) in Area B were the same as in Area A.
My system was equipped with RTK GNSS, so the camera locations are also known to a fair degree of accuracy (say 10 cm).
In order to effectively have a single model with area B more detailed than the remainder of area A, I have done the following:
1) In a single file, process each flight in a separate chunk:
1.1) Run photo alignment,
1.2) Imported the GCPs in each chunk
1.3) Adjusted each GCP in the relevant photos
1.4) Run Optimization. At this point, each chunk was georeferenced and its camera calibration optimized.
2) Workflow => Merge Chunks (I DID NOT do Workflow=> Align Chunks before this). Merged the 2 chunks; the merged chunk correctly had 2 camera calibrations, one per set of photos, as I wanted.
3) The GCPs in the merged chunk were doubled, i.e. in the Reference => Markers Pane, I had 2 Markers named 001, 2 Markers named 002, and so on; per each pair, one Marker provides errors for photos in Chunk A, the other Marker shows errors for photos in Chunk B. On each picture containing a GCP, I had two flags with the same Marker identifier (e.g. 001). I did not touch them, but I am sure that the GCPs (Markers) should be merged somehow.
4) Run Optimize Cameras on the merged chunk: for each pair of Markers, errors decreased considerably for each Marker.
5) Run Workflow => Build Dense Cloud; Build Mesh; Build Texture; Build Orthomosaic.
Questions:
1) Is my workflow correct to get the most out of the higher resolution set of photos B?
2) Is my workflow correct to get the most out of the GNSS RTK data for camera locations?
3) How and where in my workflow should I merge GCPs (Markers), if that should be done?
Thanks a lot and cheers!
FT