Height problems between coordinates in photos from m300 and Phantom 4 pro v2

Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
Hello guys, does anyone have an experience with the height difference in coordinates between m300 and Phantom 4 pro? We have automaticly flight mission with UGCS and m300 for orthos and grids. We use the phantom 4 pro v2 for more detailed pictures for 3D reconstructions. But the photos from m300 has always difference in height more than 120 meters against the phantom 4 pro and the result cant link together in agisoft. Because the m300 photos are in correct positions, but the Phantom photos are somewhere under ground. Does anyone have some idea how to fix it? Or what should we change in our workflow? Thx for you reply.
 
Joined
Aug 7, 2016
Messages
1,513
Reaction score
927
Does any of your drones have an option to toggle between AGL and ASL calculating the hight?
Mybe this could be the reason.
Such a difference is not possible if the hardware is OK.
 

Meta4

Moderator
Premium Pilot
Joined
Jan 8, 2014
Messages
15,170
Reaction score
9,105
the photos from m300 has always difference in height more than 120 meters against the phantom 4 pro and the result cant link together in agisoft.
I haven't seen photos from the M300.
Can you post one image from each drone and I'll look into their metadata to see what the differences are.

Does the M300 have the more accurate RTK GPS unit?

The Phantom has two heights in its image metadata.
One is the height relative to the launch point (what you see on screen) and the other is a rough (very inaccurate) height relative to sea level.
Agisoft uses the first one and ignores the second for mapping using the Phantom.
 
Joined
Feb 3, 2018
Messages
1,010
Reaction score
746
Age
59
Location
Albuquerque, New Mexico
Hello guys, does anyone have an experience with the height difference in coordinates between m300 and Phantom 4 pro? We have automaticly flight mission with UGCS and m300 for orthos and grids. We use the phantom 4 pro v2 for more detailed pictures for 3D reconstructions. But the photos from m300 has always difference in height more than 120 meters against the phantom 4 pro and the result cant link together in agisoft.
I assume that the M300 photos are accurate. 120 meters is a butt-load of inaccuracy, so it sounds like there may be an issue with your P4P's GPS system.


Because the m300 photos are in correct positions, but the Phantom photos are somewhere under ground. Does anyone have some idea how to fix it? Or what should we change in our workflow? Thx for you reply.
Long time mapper using P4P V1, and just did some test flights of the M300 (PPK with LiDAR). So, not an expert on this subject, but taking a stab.

I'll start with the assumption that you are using SRTM data for both drones to facilitate terrain awareness. If not, then that is most certainly at the root of your problem. That said....

I assume you are using the PPK system with your M300. The only way to achieve that same level of accuracy with a P4P is to tie the P4P photos to GCP's. If you are NOT tying P4P photos to GCP's in post, you're dealing with apples and oranges here. The P4P's consumer level GPS/GLONASS is VERY inaccurate. My business partner does the post work on the photos, but it's not uncommon to see photos located all over the place in the point cloud. It's not until they are tied to GCP's that they become accurate. It sounds like what is happening is that your very inaccurate P4P metadata is being juxtaposed against the very accurate M300 (PPK) data. Based on the information you have given, I'm not surprised at your result.

Just my 2 cents. Forgive me if I'm wildly off base.

D
 
Joined
Apr 23, 2020
Messages
17
Reaction score
1
Age
49
Hello guys, does anyone have an experience with the height difference in coordinates between m300 and Phantom 4 pro? We have automaticly flight mission with UGCS and m300 for orthos and grids. We use the phantom 4 pro v2 for more detailed pictures for 3D reconstructions. But the photos from m300 has always difference in height more than 120 meters against the phantom 4 pro and the result cant link together in agisoft. Because the m300 photos are in correct positions, but the Phantom photos are somewhere under ground. Does anyone have some idea how to fix it? Or what should we change in our workflow? Thx for you reply.
Can you share a full image with metadata as well as the flight height you set in the planning parameters?


When I do not use GCPs or RTK here is how I process in Agisoft Metashape with my Phantom 4 Pro V2


When I go to process I use an online map that can give me the ellipsoid height of the first image taken.



Next I import into Agisoft, with Agisoft expecting my images to have altitudes reported as above Ellipsoid.

I then have this with altitudes coming from the Absolute Altitude field in metadata.

1661625091502.png


Now I take the coordinates of the first image taken and plug the coordinates into the map with an above ellipsoid layer and I get that the ellipsoid height on the ground of where the first image taken is -33.54m.

1661625788118.png




I then take the Relative Altitude from the first image taken from the metadata which was 56.2 meters. This altitude is from the drones barometer. I take the difference in the Ellipsoid height on the ground which was -33.54 and add 56.2 m to it to get 22.06m which now represents that drone's altitude when the image was taken above the ellipsoid.
1661626592757.png



I now get the difference from the elevation that was imported into Agisoft (Refer to the Agisoft images above where image 001 has 31.53m Altitude). This difference is 8.87 m
In Column H I then get the difference for each image's Altitude by subtracting 8.87m from the Altitudes reported. (Column H)

I will then import column H as the Elevations for my images.

You could also just use the 22.06m for all of your elevations and just ensure that you tell the software that these are inaccurate elevations. (See First Screen shot of Agisoft where I tell the software that accuracy can be up to 10 meters off).

Using this workflow will reduce my elevation error in non-GCP/RTK maps to 1 to 4 meters and sometimes sub meter.


There is a ton of debate on elevations and DJI drones when not using RTK and I have received conflicting information from DJI themselves. I made my own workflow from these discussions and found this to work quite well. Using the above my Parrot Anafi consistently will have RMSE errors sub meter when mapping.

Pix4D Support also chimed in to help me figure out if my different drones were reporting in as above the ellipsoid or sea level and just when we thought we had figured it out, a dataset made us question it again. I find DJI drones to report their elevations almost erratically and inaccurately as flying in the same area and height could have shifts that I could not understand sometimes. But my workflow above works very well for me and is not too much added work.

I will get back to you on the Matrice 300 RTK. I saw that its elevations were off by 10 meters while not using RTK, but I didn't bother to investigate since the Wide Camera's outputs were very poor. I will try to get the Zoom camera to work for mapping and maybe this will clear up.
 
Joined
Apr 23, 2020
Messages
17
Reaction score
1
Age
49
To add some more information to this discussion.

Here is metadata from an image from the above workflow taken from a Phantom 4 Pro V2.

This flight was conducted at a commanded 180 feet (54.86m) relative altitude.

Absolute Altitude: 31.53m
GPS Altitude: 31.5m
Relative Altitude: 56.2m

The height above ellipsoid (Assuming it is WGS84 ellipsoid since the metadata of the map is not saying if it is WGS84 or GRS 80) for the coordinates of the image is -33.54m. This is from an Arc Online Ellipsoid layered map.

The orthometric elevation of this point using the EGM96 gravity model/geoid is 2m which matches up well when using NAD83(2011) and hybrid Geoid 18 (Most current geoid before the new horizontal and vertical datums are released very soon). Using Hybrid Geoid 18 I get around 1.8m for my orthometric elevation.

So lets try to figure out if the absolute Altitude reported by the drone is above the orthometric height. 54.86m as commanded from the planning software or 56.2m from the Relative Altitude above either the WGS84/EGM96 height (2m) or the NAD 83 (2011) Geoid 18 height of 1.8m would be totally incorrect with these heights adding to the originals. So unless I am missing something, Absolute Altitude is not being reported as above an orthometric height.

Next let's see if this is an above ellipsoid height.
The ellipsoid height for the coordinates from this image as reported by the ellipsoid map I use from Arc Online has an elevation of -33.54m. If I take the Relative Altitude of 56.2m from the metadata and get the difference from the ellipsoid height I get 22.66 meters above the ellipsoid. If I even take the commanded from flight planning software height of 54.86m and do the same I get 21.32m above the ellipsoid. Neither of these match up well with our 31.53m, but they do match better than our above orthometric height value of around 58m.

Just guessing I would guess that 31.53 is reporting as above ellipsoid and is just very inaccurate.

A final piece of the puzzle was thrown at me while playing around with our new Matrice 300RTK and using the cool and free for now Beta of DJI Flight Hub 2.

For the same area (I eyeballed it) DJI is saying that the orthometric height is 7.6m??? and the ellipsoid height is -27.9m????. I know this area pretty well and, no the orthometric height is not 7 meters. This could be a misprint as 7.6 feet would be close to being true with it converting to 2.3 meters. Next their ellipsoid elevation does not match the Arc Online ellipsoid map (assumed to be WGS 84).

Lets now use DJI sourced ellipsoid information (The only DJI info I ever found) with our metadata.
Take the difference between the ellipsoid height from what DJI Flight Hub 2 says it is at -27.9m and apply of image's Relative Altitude of 56.2m and we get 28.3 meters above the ellipsoid for our image. This is much closer to the Absolute Altitude of 31.53m but still no cigar.


What is the point of all of this? Nothing but to get a discussion going and for me personally to see if I am missing something in my thoughts on the vexing DJI elevations.

If you see my thinking is off or have something to add please do so.
 
Joined
Apr 23, 2020
Messages
17
Reaction score
1
Age
49
I'm not the OP.
I am just trying to help the OP. I have my own system that works for the rare times I don't use RTK or GCPs.
I don't have access to my work drone files at home and I have weekends off.
 
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
Hey guys, thank you all for your replies. Here are the photos from Ph4 and M300 approximatly the same place.

Thanks
here is the link to M300 photo, because its too large to import here: DJI_20220818124108_0412-m300.jpg
 

Attachments

  • DJI_0003-Phantom.jpg
    DJI_0003-Phantom.jpg
    7 MB · Views: 18
Last edited:
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
I haven't seen photos from the M300.
Can you post one image from each drone and I'll look into their metadata to see what the differences are.

Does the M300 have the more accurate RTK GPS unit?

The Phantom has two heights in its image metadata.
One is the height relative to the launch point (what you see on screen) and the other is a rough (very inaccurate) height relative to sea level.
Agisoft uses the first one and ignores the second for mapping using the Phantom.
Hey guys, thank you all for your replies. Here are the photos from Ph4 and M300 approximatly the same place.

Thanks
here is the link to M300 photo, because its too large to import here: DJI_20220818124108_0412-m300.jpg
 

Attachments

  • DJI_0003-Phantom.jpg
    DJI_0003-Phantom.jpg
    7 MB · Views: 18
Last edited:
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
Can you post one image from each drone and I'll look into their metadata to see what the differences are.
Hey guys, thank you all for your replies. Here are the photos from Ph4 and M300 approximatly the same place.

Thanks
here is the link to M300 photo, because its too large to import here: DJI_20220818124108_0412-m300.jpg
 

Attachments

  • DJI_0003-Phantom.jpg
    DJI_0003-Phantom.jpg
    7 MB · Views: 18

Meta4

Moderator
Premium Pilot
Joined
Jan 8, 2014
Messages
15,170
Reaction score
9,105
But the photos from m300 has always difference in height more than 120 meters against the phantom 4 pro and the result cant link together in agisoft. Because the m300 photos are in correct positions, but the Phantom photos are somewhere under ground.
There are two altitude numbers in the metadata for each image file.
The Relative Altitude is the number you see on screen while flying.
Absolute height is the inaccurate approximation of height above sea level.
Agisoft only uses the relative height and completely ignores the absolute height numbers.

The metadata for those two files shows:
P4 pro
Absolute height 31.5
Relative height 56.2

M300
Absolute height 344.5
Relative height 104.2

That's a difference of approx 50 metres (not 120 m).
Were the drones flying at 56 m (P4 pro) and 104 m (M300), like the data indicates?

If you were to fly the drones at the same height, does Agisoft have any problem working with the image files?
 
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
There are two altitude numbers in the metadata for each image file.
The Relative Altitude is the number you see on screen while flying.
Absolute height is the inaccurate approximation of height above sea level.
Agisoft only uses the relative height and completely ignores the absolute height numbers.

The metadata for those two files shows:
P4 pro
Absolute height 31.5
Relative height 56.2

M300
Absolute height 344.5
Relative height 104.2

Were the drones flying at 56 m (Phantom) and 104 m (M300)?
The phantom photos were in cca 31 meters and the m300 were in 105 meters
 

Meta4

Moderator
Premium Pilot
Joined
Jan 8, 2014
Messages
15,170
Reaction score
9,105
The phantom photos were in cca 31 meters and the m300 were in 105 meters

Its weird that the right height in m300 is the relative and the right height in p4p is absolut. Any idea how to fix it?
How can you be sure that the Phantom was flying at 31 metres higher than the launch point and not 56 metres?
I've never seen the metadata mixed up.
There's no way you can "fix" it.
 
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
How can you be sure that the Phantom was flying at 31 metres higher than the launch point and not 56 metres?
I've never seen the metadata mixed up.
There's no way you can "fix" it.
Mainly because its manual flight above a flat ground and the size of the building is cca 30 meters.
 
Joined
Aug 26, 2022
Messages
8
Reaction score
0
Age
27
There are two altitude numbers in the metadata for each image file.
The Relative Altitude is the number you see on screen while flying.
Absolute height is the inaccurate approximation of height above sea level.
Agisoft only uses the relative height and completely ignores the absolute height numbers.

The metadata for those two files shows:
P4 pro
Absolute height 31.5
Relative height 56.2

M300
Absolute height 344.5
Relative height 104.2

That's a difference of approx 50 metres (not 120 m).
Were the drones flying at 56 m (P4 pro) and 104 m (M300), like the data indicates?

If you were to fly the drones at the same height, does Agisoft have any problem working with the image files?
When I open the exif of the ohantom photo, the absolute height is 218.71 and relative 28.4. how did you find the 31.5
 

Recent Posts

Members online

Forum statistics

Threads
142,310
Messages
1,462,040
Members
104,074
Latest member
Fabitamag