Skip to main content
Solved

simplifying a pointcloud by 3dgrid


koenvdw
Contributor
Forum|alt.badge.img+6

Hi 

I have a point cloud that I want to simplify. I need 1 point in every 1cmx1cmx1cm space/voxel.

 

I already did something similar, but that was in 2D with point clouds containing only ground points and with the resulting point cloud as some sort of DTM.

In short, I used the 2dgridaccumulator to make a grid. Then I calculated the mean Z value’s of al points in a grid tile and added that value to the centerpoint of that grid. Than I used the vertexcreator to make the new point cloud. 

 

Now I need to do ‘kind of’ the same thing but in a 3D grid…

I was looking at some transformers: pointcloudduplicateremover, pointcloudsimplifier, pointcloudspatialthinner, 3dboundingboxreplacer…

 

I haven’t found a really good way to do this. Anyone with ideas?

 

 

 

 

Best answer by takashi

Hi ​@koenvdw ,

PointCloudSpatialThinner from FME Hub might help you.

View original
Did this help you find an answer to your question?

2 replies

takashi
Evangelist
  • Best Answer
  • June 4, 2025

koenvdw
Contributor
Forum|alt.badge.img+6
  • Author
  • Contributor
  • June 5, 2025

Yes, the spatialthinner looks like a viable solution.

I also tried the pointcloudsimplifier, but it’s so heavy on my machine that it crashes FME. 

The input is a 7GB point cloud with around 365 million points. 

My machine is a i9 11900 with 8 cores and 128GB ram and nvidia rtx A4000. So it’s no slug, but it couldn’t handle it. Even weirder is that after the crash my cpu was kept at 100% while I couldn’t see FME as a process anymore.


Cookie policy

We use cookies to enhance and personalize your experience. If you accept you agree to our full cookie policy. Learn more about our cookies.

 
Cookie settings