CloudCompare forum

So I made a pointcloud in agisoft metashape from drone images of an area with a lot of termite mounts. I'm trying to find an efficient way to calculate all of the mounts.

This is my current method:
Tools  Projection  Rasterize: -> Mesh
Tools  Distances  Cloud2Mesh
Edit  Scalar field  Filter by Value: threshold is the lowest value which is not part of the topografic area.
Tools  Segmentation  Label Connected Components: Here I get a lot of components.
Tools  Volume  Compute 2.5D Volume

So this was easy when there was just 1 mount, when I try this with the full data it gets a little bit trickier.
At 'Label Connected Components' some mounts come with lots of extra area, but setting the treshold higher will make other mounts too small (where the treshold should be lower)

If you think of a better way to calculate this please let me know. If you have questions or would like to try yourself, feel free to ask and i will send you the data set.

Kind regards.

Have you tried the CSF plugin maybe? To extract the ground?

daniel wrote: ↑Sat May 22, 2021 9:32 am Have you tried the CSF plugin maybe? To extract the ground?

Hi Daniel,
Sorry for this late answer, I am a student and had some exams.

I tried the CSF plugin but came to a similar conclusion. The result is roughly the same (slightly worse even) see screenshots
Also this is done with a really small portion of the dataset, it becomes way more difficult with the bigger dataset (9gb). That is why i'm maybe looking for a complete different method.

The first image is with my first discriped method, the second is with the CSF plugin.


https://imgur.com/bvGqlZc
https://imgur.com/Akh3a3w

Wouldn't it be easier to segment them quickly with the scissors tool? I thought that the mounds were not separated (or are they, and you only shared images after segmentation maybe?)

With a small amount of mounts , it is indeed easier to use the scissors tool. but for the entire dataset (around 500 termitemounts or even more) it does'nt seem like a very efficient way.