CloudCompare forum

I would guess that there's an issue with how the normals are oriented. How have you computed them?

You simply have to check the 'Visible' checkbox in the options:

Yeah, Canupo should work great. What parameters are you using? (maybe you could share a screenshot). Have you used very/too large scales maybe?

Well, I bet that sadly this is a closed proprietary format... I don't have any information about this format.

Yes, that would help a lot! Thanks.

Hum, there's a command to batch export this kind of information for point clouds... But not sub-meshes (nor meshes) sadly.

If you have C++ development skills, I bet it wouldn't be too hard to replicate for (sub)meshes.

It looks like there is a scalar field (you can activate/display the color scale in the cloud properties). But the cloud is very sparse. Maybe you can reduce the default subsampling step for Core points generation (or simply use the original points if your cloud is not that big).

Maybe you are selecting the folder including the mesh and not the mesh itself? Or the vertices and not the mesh?

So CloudCompy is a different project (here I was suggesting to use 'Python Runtime'). For CloudCompy, you'll have to follow the instruction on the project website or on its github page: https://github.com/CloudCompare/CloudComPy/blob/master/README.md It's also more efficient to ask questions on both...

I'm not the expert, but I believe as long as the other tiles have the same density and objects of the same scale, then the classifier should work on these tiles as well.

And you can normally automate the process via the command line.