CloudCompare forum

Hello!

I am looking for a way to find the volumetric difference between two LiDAR point clouds of cave data: One includes bats inhabiting the cave, the other does not. The goal is to use the calculated volume difference to thus calculate a bat census. The idea is to use the difference in addition to an estimated individual bat volume to create our rough estimate. The bats are quite 2small and there are hundreds of them; this has pushed me away from utilizing the M3C2 plugin to individually identify the figures.

I am quite new to working with point clouds, so I always appreciate a nice over-explanation; I appreciate even simple suggestions!

Thanks, S

Can you maybe show us what the cloud with bats looks like? It would help us guess if it's 'super hard' or 'impossible' ;-)

hi Daniel - apologies for the delayed response. Below is one segment of the cave.

I clipped the point cloud to only the roof of the cave for ease of interpretation. Imagine you are standing inside a cave, looking straight up at the ceiling.


*please note that this data is not owned by me individually and is not meant for public circulation*

I would be happy to supply more images or context if needed. Finding a way to isolate the bats from these point clouds (it does not matter which route to get there) is main goal of my project, so I want to do everything I can, even if it's "super hard":) Looking forward to your input!

Ah, if it's on the ceiling, maybe the 2.5D volume calculation tool would do the trick?

You would have to segment both clouds (with and without bats) the same way, then maybe manually rotate them so that the ceiling is as horizonal as possible. Then use the 'Tools > Volume > Compute 2.5D Volume' tool, choose an appropriate grid size, and see what happens ;)

If you can send me the clouds, I can also try to see what can be done on my side (admin [at] cloudcompare.org). I never disclose the data I receive 0:-)

I wonder if qCanupo would help in the segmentation, the bats definitely seem to have a distinct shape that looks like it would be classifiable.