Use thermal Images for texturing of 3D Model

Hallo everyone,

i am using a Mavic 2 Enterprise Dual to take pictures of a building. My Goal is to use the RGB pictures to create the 3DMode/pointcloudl of the Building(which works well) and then use the images from the IR sensor for textureing.
I also tried to create a pointcloud directly from the thermal images but the resolution is very bad and since the drone takes rgb and thermal images at the same time, it seems unintuitive to not use this information.
Is this workflow possible with WebODM, or does anyone have an idea, how to do this?

1 Like

I suspect that it might be very difficult due to the low resolution of the IR images cf the visible light images. Also, bright points in IR wont generally correspond with bright points in visible light.

My only suggestion, which probably wont work, would be to re-sample the IR images to the same resolution as the visible images (they will look horrible at full size!) and greatly reduce the number of features and mesh size.

1 Like


Good use-case.

Unfortunately, I don’t believe we currently support using arbitrary bandsets for reconstructon vs texturing.

Ok, thank you for your answer. I assume it is not easy, to stop the Process before texturing and switch out the pictures but Safe the referenceinformation?

1 Like

I’ve not tried it, and I don’t know if others have, but you may be able to cancel just before the texturing phase, swap the images, and then do a --rerun-from mvs_texturing

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.