Skip to main content

I want to create boundaries on some point cloud data (see https://knowledge.safe.com/questions/109029/how-to-create-disjoint-boundaries-of-disjoint-clus.html )

 

Note that the point cloud data were previously stored in an oracle dabatse as multipoint

 

For some reason, the ETL get stuck during the boundary polygons creation with the HullReplacer

 

My datasets is very big (~ 26 millions of points). I would like to know if there are some transformers like pointcloudthinner for single multipoint in order to reduce the density of the input points or if there a reverse transformer for pointcloudcoercer so that I can reconvert first the single multipoints to a pointcloud and then use for instance the PointCloudThinner before sending the data to a Hull replacer

 

Thanks

@arthy how big is your dataset? And it looks like you are clipping the dataset!


Hi @arthy,

Have you already taken a look at this article on Creating Boundary and Point Features from a Point Cloud? It uses the HullReplacer to create boundary polygons from point cloud input and might be a good reference for what you're trying to do. There's also a completed workspace in the example that you can download and explore. Hope that helps a bit!


@arthy how big is your dataset? And it looks like you are clipping the dataset!

@srg

Approximately 26 millions of single points.

Yes I'm using a clipper. However, when I disable the HullReplacer, I have the results I want but the moment I enable it, it get stuck


Hi @arthy,

Have you already taken a look at this article on Creating Boundary and Point Features from a Point Cloud? It uses the HullReplacer to create boundary polygons from point cloud input and might be a good reference for what you're trying to do. There's also a completed workspace in the example that you can download and explore. Hope that helps a bit!

Yes @jovitaatsafe.

 

I had and haven't found an answer to what I'm looking for.

Yes @jovitaatsafe.

 

I had and haven't found an answer to what I'm looking for.

Hi @arthy,

For thinning a point cloud, it's probably better to do it before you coerce it to a multipoint so you don't have to convert back forth between point cloud and multipoint. However, if you do really need to go between, I think the PointCloudCombiner might do it.

When features enter the HullReplacer, are these already coerced to multipoint? I believe the HullReplacer currently only takes in multipoint features.

If thinning the point cloud first doesn't work for you, are you able to share your workspace and a small sample dataset?


Reply