• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

oversized buildings

dave hoeffgen

dave hoeffgen

Resource contributor
Messages
1,439
Country
germany
Hi,
trying to make autogen for a larger area using openstreetmap data I got a problem.


I used the following code part for buildings:
Code: 
AddAttribute|FROMFILE="buildings.shp"|String;Building|Yes
UnloadFeatures|FAREA>1000
CreateAGNGenBuild|Building="Yes" AND FAREA<1000|{5ae04eb6-934c-4f63-bb48-5e7dee601212}|MAXRATIO=3

My problem is that despite my redundant attempt to set an upper limit for the area of polygons used for building creation I end up with some huge buildings in my scene, some of them even filling entire grid cells :eek:

Is there something I missed?
 
Hi,

You are processing all buildings, that's your problem. Autogen is always made for the rectangle that fits around the polygon. So if the polygon has a weird shape the building will turn up too big.

What I typically do is check the shape of the building as well. I believe there is an example in the manual as well. But an easy way to get only buildings that are close to rectangular is to add FAREARAT>0.8 to your filter. This will only process buildings where the polygon covers at least 80% of the bounding rectangle.
 
arno said:
You are processing all buildings, that's your problem
Click to expand...
Actually what I checked the data and there are often no buildings at all at the places in question.

Anyway I'm currently trying the filter you recommended and will report back in a moment.
 
I must say I am confused. :confused:

I double checked the data and did not encounter any polygons in the areas where that happened. Also there are no reallsy weird shapes included, almost everything is rectangular.
But still the filter solved the issue and the "giants" are gone.

Thanks for the help
 
Arno,

Could his issue be the same I experienced whereby a building (or a string of buildings), is created matching an entire LOD13 border? Just sounds similar and I thought you mentioned there was some filtering issue you needed to work on. This was before the holidays.

Dave... a suggestion. As a work around for me in my statement I made sure no building was larger than say 400 meters in Width. That filtered out the LOD13 size buildings. Maybe worth a try if it's the same issue I had.
 
Clutch Cargo said:
Dave... a suggestion. As a work around for me in my statement I made sure no building was larger than say 400 meters in Width. That filtered out the LOD13 size buildings. Maybe worth a try if it's the same issue I had.
Click to expand...
As I said it happened in places where there were no buildings at all in my shapefile, meanwhile the ratio filter already helped.
Thanks anyway for the hint.
 
Hi,

Having polygons cover an entire LOD tile would be weird. If it's a SplitGrid bug it might help to save the output after splitting to file and see if such big polygons have appeared. But in most cases it's about an incorrect filter that results in more polygons being processed than intended.
 
Hello again,
sady the "giants" have reappeared, no clue why.
I called an unloadfeature command with both the farearat and fwidth filter with no effect.

Indeed some of the buildings match the grid in annotator (see screenshot below), meanwhile most are smaller.

I repeatedly checked my data source and it does not contain any features in the places in question.

Any idea?
 

Attachments

  • giants.jpg
    giants.jpg
    929.8 KB · Views: 374
Could you send me a sample of the data so I can try to reproduce ?
 
Thanks, I'll try to have a look later today.
 
Hi,

Hmm, that's a big amount of data. With a sample of the data I meant a small big that shows the problem :).

This amount of data is so huge, that I can't easily debug it either.

One thing I notice from the script is that you split your buildings. I never do that, since that will result you in two smaller buildings where they cross a LOD tile. So I typically specify the buildings as features that should only be sorted in the LOD tiles, not split. The line would become then:

SplitGrid|AGN|*|FROMFILE="buildings.shp"

Maybe that solves your problem. Although it sounds like a bug in SplitGrid with certain features. But I can't test that with such a big file and without a photo scenery to view the resulting AGN on.
 
OK I tried to except the buildings from the split but it didn't help.
I will extract an area of the shapefile once I get the time. Shall I send you the resulting agn files too?

What also made me suspisious was that on building creation (currently the only autogen step) the report says less features were created than exported later:
15:28 CreateAGNGenBuild Information Created 624897 autogen objects
15:28 ExportAGN Information Created 625049 generic buildings objects

Maybe that helps to find the issue.
 
If you have an extract of the input data that should be enough. I can't view AGN files without a photo scenery anyway.

Those counts are sometimes a bit weird, I have seen that before. I think that might be due to the parallel processing.
 
Hi,

That's easier to process. I had a look but didn't find a bug in the SplitGrid step. The polygons after splitting don't show weird polygons that fill a big part of the tile. So the problem should be in step that makes the buildings.

If you can narrow the area down to just one tile that has the problem I can probably test it without having a photo scenery that matches.
 
Hi Dave,

I have fixed a bug with the counts. When incrementing the counter in different threads that was not always save. So that's why you saw the different numbers. But it should not affect the resulting scenery.
 
Sorry for the silence,
here is another subsample of the shapefile just somewhere around one of the affected cells.
 

Attachments

Thanks, I'll try to have a look at it and see if I can reproduce it.
 
Hi,

I had a look at the last sample file, but I can't reproduce the issue. If I look at the AGN files (I don't have a photo scenery, so can't check in Annotator) I don't see big buildings in there. Are you sure you see it with the sample files?
 
You must log in or register to reply here.
Back
Top