generating a multispectral orthophoto with Mavic 3 multispectral

[Timberdrone]

I'm attempting to generate an orthophoto using my Mavic 3 multispectral. 
In reading through other posts, I found this:

"to build a multispectral orthomosaic you first need to create a multispectral project and you have two options for that:
just use "Add Photos" and choose "Create multispectral cameras from files as bands" or from layer
if you don't have the options mentioned above, you have to put each band into a separate folder (make sure to have each index for all bands), use "Add Folder", choose the parent folder of your band folders and choose "Create multispectral cameras from folder as bands"

I have attempted to add the photos as a group, but did not see the "Create multispectral cameras from files as bands" in the workspace palette.  So, I arranged the photos into subfolders, sorted by band and attempted to add them using the "Add folder" option, selecting the parent folder that contains the subfolders for each band.  Again, I did not get the "Create multispectral cameras from folder as bands" option in the workspace palette.  I only see "arrange images based on meta data" and "create sensor from each sub folder."  Selecting either of those options only loads the images from the RGB camera, and ignores the additional bands.  What step am I missing? 

[DoubleJ]

I don't see  "Create multispectral cameras from files as bands" anywhere but I'm also having problems loading photos.

If I choose "Add Folder" it will only load the 'green' bands.

If I choose "Add Photos" then filter only '*.tif" images or even put all the tif images in a separate folder, I can only load all monochrome bands if I choose "Single cameras". If I select "Multi-camera System" then only the 'green' bands will load into the chunk.

I don't think Metashape is functioning properly for Mavic3 multispectral unless you are only trying to process the Bayer array RGB captures.

[olihar]

This is not supported with Metashape as you are trying it, you need to build custom pipeline in front of metashape for this to work.

[jmgc]

This is not supported with Metashape as you are trying it, you need to build custom pipeline in front of metashape for this to work.

Hi olihar, what kind of custom pipeline? RGB Band split, co-registration with multispectral data, to be able to process all together?

My expectation is that Metashape will have a proper worflow for Mavic 3M images processing.

[mbaus]

Hi there,

I just thought I'd post that we are processing our M3M data in Metashape just fine.

Key steps:
- Add Photos, and select only the .tif files
- Load as a multicamera system. This will appear that only the _G files have loaded (green band), but I promise they are all in there
- Run alignment
- We treat the point cloud densification as optional - we don't care about the DSM, and skipping densification saves heaps of time
- Create DSM
- Calibrate Reflectance (use spectral panels if you have them, otherwise just tick the box for the sun sensor)
- Mosaic with "average" blending - we get stripes in the mosaic if we use the other options. I've never found an official reason for it, but my guess is that its due to the pitch of the drone as it points towards and away from the sun with North-South transects
- Don't forget to convert to reflectance afterwards (we do this in QGIS as part of our QAQC process)

Hopefully that helps someone!

[jollyb]

Key steps:
- Add Photos, and select only the .tif files
- Load as a multicamera system. This will appear that only the _G files have loaded (green band), but I promise they are all in there
- Run alignment
- We treat the point cloud densification as optional - we don't care about the DSM, and skipping densification saves heaps of time
- Create DSM
- Calibrate Reflectance (use spectral panels if you have them, otherwise just tick the box for the sun sensor)
- Mosaic with "average" blending - we get stripes in the mosaic if we use the other options. I've never found an official reason for it, but my guess is that its due to the pitch of the drone as it points towards and away from the sun with North-South transects
- Don't forget to convert to reflectance afterwards (we do this in QGIS as part of our QAQC process)

Hopefully that helps someone!

This was super helpful and just in time for some field work, thanks a lot for posting!

[March Castle]

Hi there,

I just thought I'd post that we are processing our M3M data in Metashape just fine.

Key steps:
- Add Photos, and select only the .tif files
- Load as a multicamera system. This will appear that only the _G files have loaded (green band), but I promise they are all in there
- Run alignment
- We treat the point cloud densification as optional - we don't care about the DSM, and skipping densification saves heaps of time
- Create DSM
- Calibrate Reflectance (use spectral panels if you have them, otherwise just tick the box for the sun sensor)
- Mosaic with "average" blending - we get stripes in the mosaic if we use the other options. I've never found an official reason for it, but my guess is that its due to the pitch of the drone as it points towards and away from the sun with North-South transects
- Don't forget to convert to reflectance afterwards (we do this in QGIS as part of our QAQC process)

Hopefully that helps someone!

Thanks for this guidance.

Has anyone characterised the sensors of the Mavic 3M? I am trying to convert the D.Ns into reflectance and will need to perform some cross-calibration with another instrument unless other users have found an alternative method.

[xyzm]

Getting back to this old thread, if someone has experienced similar things with Mavic 3M

I am trying to create reflectance-calibrated multispectral ortho, but it seems that using reflectance calibration either with sun sensor or targets does not work correctly, as it makes half of the created orthomosaic dark. RGB is fine, as well as creating a ms mosaic without any calibration

Any help 

[olihar]

Can you upload the data somewhere, it is little hard to help without seeing the data.

You can send me a link in a message if you dont want to share dataset publicly.

[xyzm]

I have looked into this issue in more detail, and now I think that reflectance calibration works differently for the images taken before or after the target image. I took photos of the targets mid-flight. Before reflectance calibration colors look good, but after calibration, images taken before targets turn very dark, and images after calibration seem more light. Please let me know if you have any suggestions

[xyzm]

Can you upload the data somewhere, it is little hard to help without seeing the data.

You can send me a link in a message if you dont want to share dataset publicly.

I have looked into this issue in more detail, and now I think that reflectance calibration works differently for the images taken before or after the target image. I took photos of the targets mid-flight. Before reflectance calibration colors look good, but after calibration, images taken before targets turn very dark, and images after calibration seem more light. Please let me know if you have any suggestions