Merging chunks of different cameras

[Helsky]

Dear Agisoft team,

I have several projects that I have captured with two different cameras (one visible, one near infrared) at the same time / same area. I have processed them (align, DEM, Orthomosaic) in different chunks and used very high precision ground control points to georeference them. Is there an option to append the bands of the different chunks during Orthomosaic export?
What I hope for is to omit an extra step of reprojecting one orthomosaic to another with another tool.

Thanks a lot

[mikoontz]

I am also very interested in this workflow. I captured imagery with both an RGB camera and a multispectral camera during the same flight with the cameras co-mounted on the same aircraft and using similar intervalometer settings (image capture every 2 seconds for each camera). The cameras did *not* trigger at exactly the same moment, but did cover the same area under very similar lighting conditions.

I'd like to build a dense point cloud using the higher resolution RGB imagery, then link the two sets of images using manual tie points, then use the RGB-derived dense point cloud to orthorectify *both* the RGB imagery and the multispectral imagery. The final product I'm looking for is a multiband geoTIFF with pixel-by-pixel co-registration of the RGB and multispectral orthophotos. This is possible with Pix4D (see https://community.pix4d.com/t/desktop-cloud-hybrid-pix4dmapper-pro-4-3-31-imagery-with-different-spectral-signatures-one-block-after-subproject-merge-two-blocks-after-cloud-processing/7773) and I'd like to get it to work with Metashape!

Helsky-- if this use case is too different from the one you are trying to get help with, I will create a new thread! Just let me know.

[Alexey Pasumansky]

Hello,

Currently in Metashape it is not possible to merge orthomosaics from different chunks into multi-layer GeoTIFF. The only way to get multispectral orthomosaic in the current release version is to use multi-camera system approach, providing that sensors are synchronized and their relative positions do not change.

[mikoontz]

Thanks for the reply Alexey. I should be more clear-- I can combine separate, single-band geoTIFF files after the fact. What would be key for the Metashape processing is that those geoTIFFs are co-registered on a pixel-by-pixel basis. One way to achieve this might be to link the multispectral dataset with the RGB dataset by hand-identifying points in the photos that correspond to the same object on the ground, and then to use the dense point cloud from one dataset as the surface in the "Build orthomosaic..." step. Is this also not possible?