• 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.

P3D v5 Possible to detect size of water tank via texture filter?

crushnik

crushnik

Messages
539
Country
us-missouri
Is it possible to use points from OpenStreetMap data and use ScenProc's texture filter to determine area, and assign a different library object based on the area? Seems like a long-shot but I figured it's worth asking.
 
Hi,

I have never tried that. Does the OSM data not have the footprint as well? It's easier to get the radius that way.

But else maybe the clustering step in the texture filter can be used and then only select those with a point inside.
 
I found it is possible... but it is not consistent. For example, sometimes OSM and MS footprints may provide an actual footprint of a tank. From there you may be able to acquire a size (area), and then determine which library object to use based on that. But at other times you may only have a point or a point with no reference as to what that point represents. Works nice when all the pieces are in place but I have found many times I have to manually scan and place tanks by hand. Hit-n-miss.
 
Even after I've filtered out water tanks/storage tanks from the osm.pbf file it still takes FOREVER to inspect the data for a particular point so I didn't know they had size information. I should be able to inspect it on the OSM website though. If so, that would be very easy to create one library object or another depending on data already in OSM. Thanks!
 
I'm not sure you understood me. OSM does not provide size information. There actually no database available for storage or water tanks. They may provide only a point or a footprint. You would use scenProc to approximate the size of that footprint and then have the script determine what size tank to use. If it's a point, not much one can do as I see unless there was a way to match that point against photo ground imagery and come up with some size that way. Otherwise hand annotation is perhaps the only method.
 
OSM's wikipedia page on water tanks shows there is a point for the map feature, plus sometimes a polygon of the tank's outer circumference. Randomly assigning library objects (large and small) in ScenProc might be the best path forward since there are so many.
 
Did you try to cluster the image and use these features to estimate the radius? Would be interesting to try. Let me see if I have some imagery with storage tanks to test it.
 
I had to search for "cluster" in the manual. Is that part of the Terrain Filter Editor's "KMeans?" That is currently way beyond my skill level, and my plan was to randomly assign a big tank and a small tank. :) Interested to hear whether you try it and come up with anything, because it is fun coming up with new ways to model scenery. Adding tanks has definitely made flying more enjoyable.
 
I did a quick test with the K-Means clustering and the Multi Resolution Segmentation on a piece of image with storage tanks on it. But it is quite hard to capture the tanks good enough to be able to determine their radius. Often there is noise, shadows and other things that distract the segmentation. So unfortunately it's not as easy as I hoped.
 
You must log in or register to reply here.
Back
Top