aligning texture-less images based on high-precision location/orientation

[xabierr]

Is there a way to "force" alignment of images with no texture (e.g. over water) based on high-precision location/orientation information?

cheers,

Javier

[Alexey Pasumansky]

Hello Javier,

Please check the following script that allows to apply camera transform matrix according to the loaded to the Reference pane coordinates and camera orientation angles:
https://github.com/agisoft-llc/metashape-scripts/blob/master/src/quick_layout.py

Also if you have precise EO/IO information in the supported formats you can use Import Cameras command from the File menu.

[xabierr]

Thanks Alexey. I used the script, extended the bounding box size to the area of interest and tried to generate a mesh using the Extrapolated option... but got an empty surface as only a handful of tie points were present.

 

[Alexey Pasumansky]

Hello Javier,

If the overlapping images are lacking the tie points, most likely you wouldn't be able to generate the dense cloud or even build mesh from the sparse cloud. However, you can import the external mesh or DEM (for example, a rough geoid model for the area) and use the external surface to project the images onto for the orthomosaic generation.

[xabierr]

Ill give that a try thanks. How do I project the images onto the surface?