Hi all,
I cannot figure out how CC actually apply spatial reference systems. I try to be as clear as possible with this following example:
The workflow we then followed is:
Now... the model is pretty good, actually great to be used to show, but the altitudes are way off (like 30 meters, 30% off since the actual altitude is 105 meters and the model shows 130 to 135 meters) and GCPs positions are 50 cm off, which should not happen when you set ground control points to produce accurate scale, orientation and spatial position.
Could you please help us finding the issue? I was aware of a ContextCapture bug on the altitudes (which is crazy for a reality modeling software which costs a lot, the accuracy should be its job) but with so many settings for SRSs in the same production, maybe I'm a little lost.
Thanks.
It is not so much bug of CC regarding altitudes but fact that DJI P4 doesn't write GPS altitudes in EXIF, it writes barometric altitude so it doesn't matter which geoid is used but it seems more accurate after adding photos is to choose WGS84 ellipsoid(just keep in mind that it will set it as default for next imports, very strange GUI approach). But it doesn't matter if GCPs are used if they are in correct SRS. Keep in mind there are 2 SRS needed for horizontal and vertical directions. What geoid does the survey points use?
Also there is no need to mess with coordinates and trying to change before aerotriangulation. Contextcapture anyway internally uses ECEF coordinates so it recalculates them from any coordinate system to ECEF for displaying together.