CloudCompare forum

Hi Guys!

We have a .xyz data from a multibeam sonar.

How can I convert from point cloud form to 3d model form?

Thanks!

Edit--> Mesh-->Delaunay

Also

Poison Surface Reconstruction. Depends on the data, but I would say Delaunay will do fine.

Thanks a lot for your help.

These are the results.

How can I do it cleaner?

If you have any suggestions I would be glad to listen

Delaunay is usefull for more "planar" stuff like the ground seen from an aerial scaner.

In your case I would say the Poison Surface Reconstruction may be the tool you need.

https://www.cloudcompare.org/doc/wiki/i ... n_(plugin)

Also, may be interested in Meshlab
https://www.meshlab.net/

These are the outputs. Is it all I can do about them?

I mean is there any way I can clean them?

when you perform the Poison Reconstruction make sure the octree is the highest value you've since your cloud is not very dense.

If still the output is bad, I would say is the max you can do. At least that I'm aware of. May be there is a way to fit planes since your geometry is rather planar, but I'm not used to do so.

ANd I forgot of a thing, to clean up a poison mesh you can us the Segmentation Tool / the scisors icon. To select and delete the undesired faces.

As an alternative may be you could draw polylines along the "tunnel" and then use the tool that creates a mesh between polylines. I think this last one would give you a very clean low poly mesh, but not sure what you're after if super clean simple model or most accurate possible with available data.

And don't forget to output the 'density' scalar field, as explained in the dialog. This will allow you to remove the parts with low density = too few supporting points. See the wiki as well.

Hi dear friends

The helicopter blade deforms during rotation and I want to model the structure of the deformed blade.
in other words, I want to estimate the model for the deformed blade shape by using the 48 3-dimensional points. Modeling with the NURBS method had a high error because the sampling density of points was very low.

What modeling method do you suggest for my data?

ugurdemirezen wrote: ↑Tue Apr 05, 2022 8:32 am I mean is there any way I can clean them?

Use CC poisson reconstruction then clean with Blender software ...
.
Can you attach your pointscloud here in zip file?