Solved

How to filter a pointcloud based on min z from DTM


Badge +1

Hi

 

I would like to filter my pointclouds based on the minZ from a corresponding DTM.

 

So the height difference in the area's is not that big. Therefore we could use the minZ from the whole DTM. If it can be done to do it for each rastercel, I'm also interested.

 

I want to use that minZ value to filter/split/test my pointcloud.

I can't seem to find the right transformer to do this.

 

Like for the pointcloudfilter. I'm feeding it the pointcloud and from the raster(DTM) side an attribute (minZ). The filter won't accept the attribute.

 

 This is what I have right now:

 

 

fme1 

icon

Best answer by koenvdw 3 June 2021, 23:23

View original

5 replies

Userlevel 2
Badge +13

Hello @koenvdw​ , I think you may need to merge the two datasets before sending it to the PointCloudFilter, as FME won't do it automatically. You can use a FeatureMerger before the PointCloudFilter to merge the minZ from the TIFF onto the point cloud. Hope this helps - Kailin.

Badge +1

Hello @koenvdw​ , I think you may need to merge the two datasets before sending it to the PointCloudFilter, as FME won't do it automatically. You can use a FeatureMerger before the PointCloudFilter to merge the minZ from the TIFF onto the point cloud. Hope this helps - Kailin.

Hi thx for your response.

I tried the feature merger, but based on what do I join?

They don't have anything in common.

 

Spatialfilter also doesn't work.

Userlevel 2
Badge +13

Hello @koenvdw​ , anytime! Hm.. I think in the FeatureMerger, you should be able to set Requester and Supplier values to 1 for an unconditional merge. Let me know if this works - Kailin.

Badge +1

Hi @kailinatsafe​ 

 

I will certainly test that and come back to you.

In the meantime I got a tip from @Jasper Wisbecq​ to use the clipper.

This seems to work.

 

The workbench looks like this now:

fme1This is before and after:

fme2So it works pretty well.

 

Thx guys.

Badge +1

Hi @kailinatsafe​ 

 

I will certainly test that and come back to you.

In the meantime I got a tip from @Jasper Wisbecq​ to use the clipper.

This seems to work.

 

The workbench looks like this now:

fme1This is before and after:

fme2So it works pretty well.

 

Thx guys.

I tested the unconditional merge with the feature merger.

It also works and it was 15 seconds faster.

 

Also strange, there is a difference in total points and thus also in filesize.

The differences are not visible to the eye.

 

clipper method result: 4min29 - 56888763 points - 1999997KB

feature merger method result: 4min16 - 56902395 - 2000476KB

 

So there you have it, to working solutions 😃

 

 

 

Reply