Altitude used in Processing UAV (DJI Phantom 4 Pro)

[jazzyj]

Couple basic questions.  The Beginner Workflow published for DEM / Orthophoto generation lists a step to press the button to import camera positions from EXIF data.  However when I add my photos and go to the reference pane, it already shows the GPS coordinates from the EXIF data.  So is this step not necessary then because it's doing this automatically?

I've seen various posts that the barometric altimeter on the DJI Phantom 4 drifts as the drone warms up during flight.  However the reference pane is showing altitude based on GPS measurement which is good because it doesn't have the drift.  So does Photoscan thereby not use the Above Ground Level data and therefore the drift issue is mitigated since it's using GPS elevation?

[Alexey Pasumansky]

Hello jazzyj,

PhotoScan is reading GPSAltitude tag from the image EXIF and writing it's contents to the Altitude column of the camera position. And it assumes that this value represent the height above the ellipsoid. So if the base level for the loaded altitude values is different (or the values are incorrect by any other reason), PhotoScan wouldn't know that.

[ruyi7952]

This is a joke.

DJI company intentionally saved a "error" height value in Exif information!

The height shown in DJI GO App is the vertical height of the aircraft relative to the take-off point.

The two Altitude value are completely different.

I think it's time somebody came forward and sued the dji.

[jazzyj]

But as far as the accuracy of the Photoscan output, does it matter that Photoscan is using Above Sea Level (ASL) instead of Above Ground Level (AGL) data as far as determining relative elevation changes in the terrain?  That's relative elevation between points in the property, not absolute elevation ASL measurements of the terrain.

[ThomasL]

This is a joke.

DJI company intentionally saved a "error" height value in Exif information!

The height shown in DJI GO App is the vertical height of the aircraft relative to the take-off point.

The two Altitude value are completely different.

I think it's time somebody came forward and sued the dji.

I think this is more like a problem with the GPS. DJI drones are using the built-in GPS for location data and we all know it's very inaccurate, so that's why we need GCPs to calibrate the results.

[ThomasL]

But as far as the accuracy of the Photoscan output, does it matter that Photoscan is using Above Sea Level (ASL) instead of Above Ground Level (AGL) data as far as determining relative elevation changes in the terrain?  That's relative elevation between points in the property, not absolute elevation ASL measurements of the terrain.

From the 3D reconstruction point of view, I guess the relative elevation should be pretty good. I'm also kinda curious whether the GPS inaccuracy will affect the relative elevation.

[SAV]

Hi jazzyj,

The altitude EXIF issue of DJI UAV imagery is quite annoying.

Try one of these three 'workarounds' outlined in one of my previous posts:
http://www.agisoft.com/forum/index.php?topic=4986.msg38769#msg38769

I hope that helps.

Regards,
SAV

[jazzyj]

Hi jazzyj,

The altitude EXIF issue of DJI UAV imagery is quite annoying.

Try one of these three 'workarounds' outlined in one of my previous posts:
http://www.agisoft.com/forum/index.php?topic=4986.msg38769#msg38769

It may be annoying but I've yet to see someone explain why it might be significantly detrimental to the accuracy of the Photoscan output.

It is widely known the barometric altitude drifts up to 3 meters over the course of the flight as the drone heats up while flying.  So towards the end of the flight the drone might think it's at 200 feet but it's actually at 190 feet.  If Photoscan used the barometric altitude, then it would negatively effect the accuracy.

I take a photo before take off and after landing.  I confirmed that even after warming the drone up the altitude drift is 1.2m based on the GPS reading.  The GPS altitude is quite accurate for measuring change in the drone's altitude.  I had a mission at 250 feet and the different in the GPS altitude before take off and immediately after reaching flight altitude was 76 meters or 249.3 feet.  The .7 foot error is probably well within the margin of error of the barometric altimeter.  But the GPS altitude won't drift like the barometric so it's actually great that Photoscan uses the GPS altitude instead of the barometric altitude (assuming I'm not overlooking some significant drawback.) 

[Off The Ground]

Retrieving the .DAT file from the AC should help fix your issue, some what.
http://www.datfile.net/DatCon/retrieveV3Dat.html

[cadm8]

... If Photoscan used the barometric altitude, then it would negatively effect the accuracy

You're speculating on the software's performance by inspecting an individual sensor reading. Run a test project the correct way (GCPs georeferenced etc) and test it against survey data to verify your results

[jazzyj]

Retrieving the .DAT file from the AC should help fix your issue, some what.
http://www.datfile.net/DatCon/retrieveV3Dat.html

This URL generates a malware warning and my malware detector (Bitdefender) is pretty darn reliable.

[gto234]

Couple basic questions.  The Beginner Workflow published for DEM / Orthophoto generation lists a step to press the button to import camera positions from EXIF data.  However when I add my photos and go to the reference pane, it already shows the GPS coordinates from the EXIF data.  So is this step not necessary then because it's doing this automatically?

I've seen various posts that the barometric altimeter on the DJI Phantom 4 drifts as the drone warms up during flight.  However the reference pane is showing altitude based on GPS measurement which is good because it doesn't have the drift.  So does Photoscan thereby not use the Above Ground Level data and therefore the drift issue is mitigated since it's using GPS elevation?

About Beginner Workflow I noticed this also when i did my first test projects. The step to import camera positions is not needed, the software does it automatically. The step must have been left over from some previous version.
About altitude don't worry. If you care about accuracy you have to use GCPs and they will provide the correct elevations of your model.
If you have Phantom 4 pro also don't use rolling shutter correction as suggested in workflow because P4 pro has mechanical shutter.

[MIG-29]

My question is why Dronedeploy have an Altitude of 239 ft (whic actually is correct altitude conducting my flight) and when I import the same EXIF files into Agisoft Photoscan Professional, the altitude is stated approx 130 meters - almost twice the altitude?



EXIF is a standard for preserving and shareing digital video/images som why the discrepancy?

Both softwares have the exact same EXIF files...


Regards
MIG-29 (Fulcrum)

[MIG-29]

...somebody in an earlier thread wrote: "PROBLEM
The usual metadata tag that is used by these applications is the ‘GPSaltitude’ tag. Currently DJI populate this from the aircraft's GPS unit. The altitude recorded here is grossly inaccurate. We need to fix this as far as we can."

This can not be an issue or correct, since the Drone is flying the mission by the GPS, so the altitude data in the exif file should be the same as the actual GPS height which also is tranponded to the control software on pad/mobile.

Why is not the exif altitude and the drone altitude the same? and if not - why do Dronedeploy get it right?

MIG-29 (Fulcrum)

[MIG-29]

I should also mention, that I fly fixed wing sdrones and have been using photoscan for 2-3 years, and when i import the images from this drone the altitude is correct. I use this to speed up the alingemen before I use GCPs and build in local reference system...


MIG-29 (Fulcrum)