signed differences issues

[claireearlie]

Hi Daniel,
Firstly, like most people out there have said, thank you for the great software!
I am having trouble with signed differences for point cloud data in x,y,z format form a Leica ScanStation2. No problems loading the data, however when comparing cloud to cloud, for precise distance computation results, the option for signed distances is not selectable, which means the distances are only positive..the point clouds are for two different scans of the same cliff surface, so they should certainly show positive and negative differences. (I am using the latest CC version)
Any ideas why this might be?
Many thanks in advance Daniel
Claire

[daniel]

Hi,

In fact this is normal because you can't get signed distance with cloud/cloud distance.

The main reason is that by default you don't have any info about whether a point is inside or outside the surface (apart if you got normals) and due to the sparse sampling of point clouds there's other minor issues.

To get signed distances, the only way for now is to compute a mesh from the 'reference' point cloud and to perform cloud/mesh distances.

We have already talked a lot about this with Dimitri Lague on this forum (and elsewhere). It would be possible under certain circumstances or with some tricks/approximations but it's not available (yet) in CC.

Best wishes,

[claireearlie]

Thanks so much Daniel,
I have been in touch with Dimitri and am currently using (or attempting to!) his M3C2 algorithm. I had heard from another fellow user that the signed diffs cloud to cloud feature was now available in CC, I must have been mistaken
Thanks again
Claire