cloud-mesh distance calculation

[pippone92]

Dear Daniel,

I am currently working on a new task, which includes the analysis of a pregenerated mesh (.2dm) by comparing it with its original laser scan data (1x1m2). Both the mesh and the laser scan data were already provided to me like that, and this is my very first approach to CloudCompare.

My goal is somehow to come up with an idea on how to evaluate the quality of the mesh by analysing the difference in elevation between the mesh and the data set (which represents my points cloud). By following the manual, I managed to compute the distances, but I do not quite understand how these are calculated by the program and how to actually read my results. Despite the tutorial videos on YouTube, and the several discussions on this forum, I still have problems at interpreting the results. I do understand the main concept behind this operation: calculation the deviation of my reference mesh from the original point that make the cloud.. but is there anything else to it?

This is what I have got so far:

screen shot of the computed distance mesh - one data set

CC.JPG (160.64 KiB) Viewed 1393 times

Note: due to the mesh resolution, the mesh is very coarse in flat areas, hence some high deviation is legimit. However, the hydraulic constructions have been correctly computed in the mesh, thus less divergence is expected there.

Would you have suggestion on how to move on? any other paper I could read to better grasp the idea of what my results are telling me?

Thank you in advance.
Filippo

[daniel]

Yes, as you realized, your mesh is way too coarse to be used as is. You should instead sample points on the mesh (with 'Edit > Mesh > Sample points') and then use the cloud-2-cloud distance.