Question

Investigate crash in RasterDEMGenerator

  • 27 January 2022
  • 5 replies
  • 1 view
R
Rank
Badge

hi,

using FME 2020.1.0.1, I find that occasionally the RasterDEMGenerator crashes when trying to rasterize a TIN. I have isolated some of the triangles (see workbench and data in zip file).

 

The crash occurs when using the BreakLines input.

Using the "Points/Lines" input is not an option, since the resulting raster has artefacts.

 

Another FME user suggested the PointCloudCombiner as a workaround. This is a very slow option, unfortunately.

 

I hope Safe can assist in fixing or working around the crash in the RasterDemGenerator (using breaklines input).

 

Kind regards,

 

Richard.

1 Attachment

5 replies

redgeographics
Rank
Userlevel 5
Badge +25

It looks like your breaklines are very thin triangles which is throwing the RasterDemGenerator off. I've tried using a Tiler, set to a tile width of 50 in both directions. That at least makes it go through the RasterDemGenerator without errors (still warnings though).

R
Rank
Badge

It looks like your breaklines are very thin triangles which is throwing the RasterDemGenerator off. I've tried using a Tiler, set to a tile width of 50 in both directions. That at least makes it go through the RasterDemGenerator without errors (still warnings though).

Hi,

There are many thin trinagles indeed. Where did you add the Tiler (and did you change additional settings?). In the example workbench I provided, I placed the tiler between the reader and the RasterDemGenerator, connecting the Tiler output to the breaklines input. The RasterDemGenerator still failed, though.

1 Attachment
redgeographics
Rank
Userlevel 5
Badge +25

Hi,

There are many thin trinagles indeed. Where did you add the Tiler (and did you change additional settings?). In the example workbench I provided, I placed the tiler between the reader and the RasterDemGenerator, connecting the Tiler output to the breaklines input. The RasterDemGenerator still failed, though.

That's where I placed the Tiler too. It works for me (FME 2021.2) with your settings as-is. Not sure what's causing the difference in behaviour to be honest. Any chance you can update your FME?

R
Rank
Badge

Hi,

There are many thin trinagles indeed. Where did you add the Tiler (and did you change additional settings?). In the example workbench I provided, I placed the tiler between the reader and the RasterDemGenerator, connecting the Tiler output to the breaklines input. The RasterDemGenerator still failed, though.

I'll give that a try. We are planning an update in the near future, we might push that a little in case this solves the problem....

 

R
Rank
Badge

Hi,

There are many thin trinagles indeed. Where did you add the Tiler (and did you change additional settings?). In the example workbench I provided, I placed the tiler between the reader and the RasterDemGenerator, connecting the Tiler output to the breaklines input. The RasterDemGenerator still failed, though.

I tried this with 2021.2.2, and for this small subset of triangles the Tiler did the trick.

 

However, when I applied this workaround to the complete set of triangles, other issues appeared and the validation failed. The error message was a little bit more verbose, but the rejected port of the RasterDEMGenerator was not used. This does not give me an option to correct errors (or choose different paths) when running the script.

 

I will stick to the PointCloudCombiner option for now, even though this is slower.

@Hans van der Maarel​ : thanks for the tips though!

 

 

Reply


Community Stats

28,707
Posts
109,478
Replies
36,551
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.