In the aux.xml file created for each image taken there is the following tag drone-dji:FlightYawDegree="+90.00" what exactly is the range and meaning of this value.  I assmed it is the the angle of deviation of the gimbal from the centerline of the aircraft and was not associated with the flight yaw. The gimbal only has a range of motion of 180 degrees so my question is this.  What does  drone-dji:FlightYawDegree mean.  Is it the flight yaw corrected for gimbal yaw or is it just the gimble yaw?  If it is the just gimble yaw, how is it measured, is 0 centered, left, or right?
Thanks in advance.

Hi, GisGuy1. Thank you for reaching out. YAW refers to the horizontal direction, which can be east, west, south, or north. "GimbalYawDegree" refers to the direction in which the camera is aimed. "FlightYawDegree" refers to the current orientation of the aircraft, where 0 degrees represents true north, 90 degrees is east, 180 or -180 degrees is south, and -90 or 270 degrees is west.  Apart from the Mavic 2, Mavic  Air 2, and DJI Air 2S, the direction of the camera is generally consistent with that of the aircraft. Hope this helps.

DJI Gamora Posted at 3-1 00:48
Hi, GisGuy1. Thank you for reaching out. YAW refers to the horizontal direction, which can be east, west, south, or north. "GimbalYawDegree" refers to the direction in which the camera is aimed. "FlightYawDegree" refers to the current orientation of the aircraft, where 0 degrees represents true north, 90 degrees is east, 180 or -180 degrees is south, and -90 or 270 degrees is west.  Apart from the Mavic 2, Mavic  Air 2, and DJI Air 2S, the direction of the camera is generally consistent with that of the aircraft. Hope this helps.

so if the Gimbal is perfectly in line with the aircraft, the value will be 90?  If so, is 0 towards the counterclockwise or towards the clockwise?

DJI Gamora Posted at 3-1 00:48
Hi, GisGuy1. Thank you for reaching out. YAW refers to the horizontal direction, which can be east, west, south, or north. "GimbalYawDegree" refers to the direction in which the camera is aimed. "FlightYawDegree" refers to the current orientation of the aircraft, where 0 degrees represents true north, 90 degrees is east, 180 or -180 degrees is south, and -90 or 270 degrees is west.  Apart from the Mavic 2, Mavic  Air 2, and DJI Air 2S, the direction of the camera is generally consistent with that of the aircraft. Hope this helps.s

GisGuy1 Posted at 3-4 13:54
[view_image]

The infoyou've been given above is confusing.
The gimbal yaw angle is a compass heading, just like the flight yaw angle.
It's not added or subtracted from the flight yaw angle.
When you look at some actual metadata, that's the only way it could be.

if the Gimbal is perfectly in line with the aircraft, the value will be 90?
If the gimbal is perfectly in line with the aircraft, the gimbal yaw angle will be the same as the flight yaw angle.

Labroides Posted at 3-4 14:44
The infoyou've been given above is confusing.
The gimbal yaw angle is a compass heading, just like the flight yaw angle.
It's not added or subtracted from the flight yaw angle.

Thanks for the reply!  

The image on the right has a flight yaw of 122.6 and a gimble yaw of -70.80.  I don't think what I' seeing in the xml is fits your explanation (even though your explanation makes sense).  

GisGuy1 Posted at 3-4 15:34
Thanks for the reply!  

The image on the right has a flight yaw of 122.6 and a gimble yaw of -70.80.  I don't think what I' seeing in the xml is fits your explanation (even though your explanation makes sense).  

122.6 seems close but its not quite right.  -70.8 makes no sense either.

GisGuy1 Posted at 3-4 15:37
[view_image]

122.6 seems close but its not quite right.  -70.8 makes no sense either.

If that data is correct, your drone has crashed and the gimbal is wrecked.
I would tend to suspect a problem with the data.

Have you got many other images with metadata like that?
All the images I've checked, show the drone's heading and the gimbal yaw angles being reasonably close to each other.

Can you upload that image to Dropbox or similar and post a link here?

Labroides Posted at 3-4 16:07
If that data is correct, your drone has crashed and the gimbal is wrecked.
I would tend to suspect a problem with the data.

This image is part of a flight plan.  I have uploaded images all images from positions 9-14 with xml files.  This is where the aircraft was making a 180 deg turn to make its next pass.  50% of the images taken during the turn seem to have "bad data".  image 12 is the image i have been posting, image 13 seems to be good and 14 is bad again.  

https://www.dropbox.com/scl/fo/1mwo4snqh7pic25zpns7y/h?rlkey=68w7xb1z2wv2ohcy2yf64ksd2&dl=0

GisGuy1 Posted at 3-5 05:07
This image is part of a flight plan.  I have uploaded images all images from positions 9-14 with xml files.  This is where the aircraft was making a 180 deg turn to make its next pass.  50% of the images taken during the turn seem to have "bad data".  image 12 is the image i have been posting, image 13 seems to be good and 14 is bad again.  

https://www.dropbox.com/scl/fo/1mwo4snqh7pic25zpns7y/h?rlkey=68w7xb1z2wv2ohcy2yf64ksd2&dl=0

Those are some odd numbers.
I checked them with my metadata reader and the numbers are all over the place just as you said.
All the mapping that I've done was with  the Phantom 4 pro.
I checked a recent mission and all the flight/gimbal yaw data has the numbers within about 5 or 6 degrees   or less, even when the drone is cornering at the end of a row.

Last week I did a manual mapping job of a small site using a Mavic 3 pro and I checked the metadata from that and the numbers are similar to the P4 pro.
Flight and Gimbal yaw are all within 3 or 4 degrees for all shots.

I'm not sure what's going on with the numbers showing from your M3M, but it's nothing like other DJI drones and it's hard to make sense of.
I can't help wondering if the numbers are accurate at all because there's no way to make sense of them.

GisGuy1 Posted at 3-4 13:54
[view_image]

Let me confirm this with our relevant team. We'll get back to you once we receive an update.

Labroides Posted at 3-5 06:08
Those are some odd numbers.
I checked them with my metadata reader and the numbers are all over the place just as you said.
All the mapping that I've done was with  the Phantom 4 pro.

So I am working on this as a project for a friend so i do not have the drone in my hands yet. She is bringing it by tonight.  

I noticed one thing that concerns me.  The folders on the SD card that contain the images do not have these aux.xml files.  The images  are the of course, but there are only a .nav, .bin, .mrk and a .obs.  Are the aux.xml files created by some software process when moving the files from the drone to the PC?  IF this is the case, I'm wondering this software is the issue.

The MRK file mentioned above contains the drone location when the image was taken along with other data, the .bin is binary, the .nav is GNSS info and the obs seems to be some sort of tab delimited observation data.  The data to batch reference these photos does not seem to natively be there.

GisGuy1 Posted at 3-6 08:29
So I am working on this as a project for a friend so i do not have the drone in my hands yet. She is bringing it by tonight.  

I noticed one thing that concerns me.  The folders on the SD card that contain the images do not have these aux.xml files.  The images  are the of course, but there are only a .nav, .bin, .mrk and a .obs.  Are the aux.xml files created by some software process when moving the files from the drone to the PC?  IF this is the case, I'm wondering this software is the issue.

It sounds like you are talking about one of the enterprise models with RTK GNSS.
I've never used those.
None of the drones I've flown generate those files, so I can't comment on what's in them.
I just looked at the metadata in each image file, the same as any mapping program would when assembling an orthophoto mosaic.

Labroides Posted at 3-6 13:35
It sounds like you are talking about one of the enterprise models with RTK GNSS.
I've never used those.
None of the drones I've flown generate those files, so I can't comment on what's in them.

Thanks,  It never occurred to me it would be stored in the EXIF data.  That makes things so much easier. I'm now pulling the data using the PIL package for python.  There is one mystery solved.  Cant believe i over looked that.  I'm hoping the Admin's can provide some clarity to the gimbal yaw values being out of wack and im good.  Thanks!

GisGuy1 Posted at 3-4 13:54
[view_image]

Hi there. If the Gimbal is perfectly in line with the aircraft, flghtyaw=gimbalyaw. Where 0 degrees represents true north. If the angle of flight yaw is 90 degrees, then the angle of the gimbal is also 90 degrees. In a typical convention for angles and directions in aeronautics or robotics, a positive angle indicates a clockwise rotation, while a negative angle indicates a counterclockwise rotation.

GisGuy1 Posted at 3-6 14:48
Thanks,  It never occurred to me it would be stored in the EXIF data.  That makes things so much easier. I'm now pulling the data using the PIL package for python.  There is one mystery solved.  Cant believe i over looked that.  I'm hoping the Admin's can provide some clarity to the gimbal yaw values being out of wack and im good.  Thanks!

Hi,

Interesting question and DJI answer to this.

Had a look in one of my pictures taken during flight and check the yaw values.

See chart YAW1 ; Flight yaw Degree = minus 136.80 and Gimbal yaw degree minus 150.20.

136.8 = 360 - 136.8 = 223.2 compass heading
150.2 = 360 - 150.2 = 209.9 compass heading
This picture was take in hover so craft steady on a fixed heading.
So why a difference between these 2 headings as on a MINI4pro (and other DJI drones) camera = looking forward = drone heading.

During powering ON your drone few yaw channels are 'loaded' with compass heading, but not all yaw channels act the same and use the same input on a drone heading change.
A heading change in flight is 'noticed' bt the compass ofcourse and changed heading written to the Flight Yaw Degree.
The same heading change is also 'noticed' by the yaw gyro and yaw accelerometer, if all works oke than the same amount of heading change is calculated, for some reason this is seen on the Gimbal Yaw Degree.
Ofcourse if the cam gets left /right offset off looking forward than this is also seen in this Gimbal yaw value.

When the 2 heading values are getting beyond a threshold value, compass/yaw errror message FlyApp, but apparantly seen from the start situation of the 2 values.
See my chart of the 2 yaw lines, during almost the whole flight a mismatch of yaw values plus/minuns 15 - 20 degrees, no errors in the log ect.

Start of flight (see lines 0 - 1000 records id), a compass change wich was not a actual heading change but just compass heading change. (not on the other line...)
So if your Gimbal Yaw is not close to the Flight Yaw, curious to see your flightlog for this flight.

Chart Yaw3, typical flight with good equal values for bith channels.

cheers
JJB

JJB* Posted at 3-7 08:28
Hi,

Interesting question and DJI answer to this.

Below is a different flight plan and some more illustrations.  The red arrows in the Gimbal angles taken as compass headings.  The yellow are the Flight angles taken as compass headings. I dont know why the aircraft would choose to measure the compass heading in this way and not just use a positive 360 degree compass.  Makes no sense but its easily handleable programmatically.  you can also obviously see that a good number of these headings are very close to 180 degrees apart from each other, but not consistently.  In circumstances where the two headings are very close to each other, images typically georeference pretty well.  in areas where they are opposite, or off by some factor, they do not georeference well at all.  I dont know how any image processing software can mosaic this data with confidence.  I did notice there if you use the "out of the box" flight planner, there is an excessive amount of overlap on each image.  This leads me to believe they intend to search for images where these headings make no sense and throw them out of the mosaic while relying to neighboring images to fill the gap.  I really need some help from DJI to sort this out as the data makes no sense.  

The F in the label is the flight yaw and the G is the gimble yaw for each image number.


This is a table of all flight and gimble yaws from this flight plan.

pydev debugger: starting (pid: 24480)
                                                                                Flight        Gimble
DJI_20230710091812_0001_MS_NIR.TIF == -179.7 : -178.2
DJI_20230710091814_0002_MS_NIR.TIF == -179.9 : -179.5
DJI_20230710091816_0003_MS_NIR.TIF == -180.0 : 0.4
DJI_20230710091818_0004_MS_NIR.TIF == 180.0 : -179.9
DJI_20230710091820_0005_MS_NIR.TIF == -180.0 : 0.1
DJI_20230710091822_0006_MS_NIR.TIF == -180.0 : -179.9
DJI_20230710091824_0007_MS_NIR.TIF == 180.0 : -179.9
DJI_20230710091826_0008_MS_NIR.TIF == -180.0 : 0.1
DJI_20230710091828_0009_MS_NIR.TIF == -180.0 : 0.1
DJI_20230710091830_0010_MS_NIR.TIF == 179.9 : 0.1
DJI_20230710091832_0011_MS_NIR.TIF == 179.0 : 0.1
DJI_20230710091834_0012_MS_NIR.TIF == 84.2 : 102.2
DJI_20230710091836_0013_MS_NIR.TIF == 61.1 : 61.6
DJI_20230710091838_0014_MS_NIR.TIF == 40.9 : -128.6
DJI_20230710091840_0015_MS_NIR.TIF == 0.3 : -178.6
DJI_20230710091842_0016_MS_NIR.TIF == 0.0 : 0.3
DJI_20230710091844_0017_MS_NIR.TIF == 0.0 : 0.2
DJI_20230710091846_0018_MS_NIR.TIF == 0.0 : 0.2
DJI_20230710091848_0019_MS_NIR.TIF == 0.0 : -179.8
DJI_20230710091850_0020_MS_NIR.TIF == 0.0 : 0.2
DJI_20230710091852_0021_MS_NIR.TIF == 0.0 : 0.2
DJI_20230710091854_0022_MS_NIR.TIF == 5.7 : 179.8
DJI_20230710091856_0023_MS_NIR.TIF == 76.6 : -114.9
DJI_20230710091858_0024_MS_NIR.TIF == 90.0 : -90.1
DJI_20230710091900_0025_MS_NIR.TIF == 156.6 : -37.3
DJI_20230710091902_0026_MS_NIR.TIF == -179.9 : 0.2
DJI_20230710091904_0027_MS_NIR.TIF == 180.0 : 0.6
DJI_20230710091906_0028_MS_NIR.TIF == -180.0 : 0.6
DJI_20230710091908_0029_MS_NIR.TIF == 180.0 : -179.4
DJI_20230710091910_0030_MS_NIR.TIF == 180.0 : -179.4
DJI_20230710091912_0031_MS_NIR.TIF == -179.8 : 0.7
DJI_20230710091914_0032_MS_NIR.TIF == 143.9 : 160.9
DJI_20230710091916_0033_MS_NIR.TIF == 62.6 : 66.5
DJI_20230710091918_0034_MS_NIR.TIF == 60.9 : -118.4
DJI_20230710091920_0035_MS_NIR.TIF == 18.6 : -152.5
DJI_20230710091922_0036_MS_NIR.TIF == 0.0 : 0.6
DJI_20230710091924_0037_MS_NIR.TIF == 0.0 : 0.3
DJI_20230710091926_0038_MS_NIR.TIF == 0.0 : 0.3
DJI_20230710091928_0039_MS_NIR.TIF == 0.0 : 0.3
DJI_20230710091930_0040_MS_NIR.TIF == 4.7 : -179.7
DJI_20230710091933_0041_MS_NIR.TIF == 68.4 : -121.6
DJI_20230710091935_0042_MS_NIR.TIF == 75.7 : -105.8
DJI_20230710091937_0043_MS_NIR.TIF == 157.0 : -38.0
DJI_20230710091939_0044_MS_NIR.TIF == 180.0 : 0.2
DJI_20230710091941_0045_MS_NIR.TIF == 180.0 : 0.6
DJI_20230710091943_0046_MS_NIR.TIF == 179.9 : -179.5
DJI_20230710091945_0047_MS_NIR.TIF == -180.0 : 0.6
DJI_20230710091947_0048_MS_NIR.TIF == 179.9 : 0.3