Hello Emlid Community,
We have been using Emlid Studio for PPK processing for a couple of years.
-
We are using a Yuneec H520E RTK UAV.
-
We are always using a Trimble base station
-
Trimble R8s
-
Trimble R83
-
Trimble R10-2
-
We have done really well with it and can get all the images to geo correct through a PPK process with a very high relative accuracy to themselves.
-
When we compare to the ground control points or terrestrial scanning, it is very obvious that the vertical is not being corrected to a high level of absolute accuracy to the GCP points.
-
The file that comes out of the Emlid UAV for geotagging the image exposure moments is called “PPKTimestamp.txt”.
-
This file has to be converted to have the same format as a .mrk file which the Emlid Studio will accept.
-
From playing around with the file, I am pretty sure the three XYZ distances at each camera exposure to correct between the UAV antenna phase center to the center of the camera sensor (CMOS) are have the opposite sign than what the .mrk file would have.
The .mrk file is APC – CMOS = mrk file value
The “PPKTimestamp.txt” is CMOS – APC = PPKTimestamp.txt value
I determined this through trial and error and mainly supported because the PPKTimestamp.txt value for the Z is negative and in the mrk examples on line the Z value is positive and all these UAV’s have the UAV sensor above not below the camera so they should all have the same value sign.
-
I am not sure, but I think I might have found the issue
-
When using the Trimble R8 as a base, the error in vertical between PPK locations compared to their fit with GCP is smaller than with R10-2.
-
The Trimble R8 distance between the area reference point and the antenna phase center is much smaller than the R10-2
-
Could this mean that we are converting the rinex file incorrectly?
I would be interested to here your input on this. I am happy to share more information. I also posted this on the Emlid community forum to see if anyone else has solutions here.