Hello,
I need help with strange CC behaviour...can't track it down to the source of the problem. I have set of photographs done by drone with GPs positions in meta data.
See below:
CC reads it as WGS and this is what I'm getting:
same as above but converted to OSGB:
My GPS altitiude is set to Sea Level:
Now as you can see Z values are all out of the place...image was taken ca. 53m above ground, CC reads it as 0.6m, OSGB shows -53m and funny thing is that we have GCP which are like:
105,231787.347,718851.868,25.269106,231783.237,718816.159,25.060107,231762.916,718828.276,25.457108,231736.719,718805.632,30.724
So GCP and image Z value are correct but CC seems to have hard times trying to interpret them. This makes GCP totally unusable cos after AT my model is like 30m below GCP and CC fails to match GCP locations with any of the photograph...so after all I'm ending up with 3D model correctly sitting in X,Y but Z is a total abstraction. I checked the same photographs in two Cc versions 4.4.9.516 and 4.4.0.344 and results are the same...
Any idea what is wrong ??.
The height values can be way off(+/- 100m) if using drone without RTK/PPK even if LatLon/XY seems correct. You could adjust height if you have known relative altitude from altimeter.
Funny thing here is that DjI RTK was used...and this doesn't explain why CC can't apply Z values given in photo EXIF data ??
Regarding altimeter..most modern DJI drones have it on board so height values are in relation to GPS+altimeter so chances for it being +/- 100m are close to zero.
Contractor used D-RTK GNSS from DJI and said that it helps with GCP and Drone accuracy but there is no way DJI software can overwrite EXIF data from drone on board GPS unit and/or altimeter with the one from D-RTK. He mention something about 3 drone model which can do that but nothing from DJI.
>height values are in relation to GPS+altimeter so chances for it being +/- 100m are close to zero
Latest findings show that DJI writes barometric altitude(standard atmospheric model) as GPS altitude in EXIF data so that is why it is always wrong even if GPS position is accurate. Correct altitude can be seen in flight log or using DJI RTK.
Answer Verified By: Krzysztof Ciacka
Yup,
I figured it out too. We had to do AT without proper Z values and used GCP for next AT iteration to properly position our models in 3D space.
Hi Krzysztof. Did you solve this problem? I'm getting crazy to use GCPs reference on my model, heights are way off, and position is not as surveyed. We used a DJI Phantom 4 Pro and a Leica GS18T for points survey.
You have my answer on the plate..se my post above:
I had to do AT without proper Z values (Arbitraty AT) and then I used provided GCP's for next AT iteration (Positioning mode: Use control points for adjustment) to properly position our models in 3D space.