P3D v5 Water Detection idea (work-around)

[Clutch Cargo]

Ah, good eye. I corrected but still get the following error at the detection step:

11:07 AM SceneryProcessor Error System.IO.InvalidDataException: Unexpected image type Emgu.CV.Image`2[Emgu.CV.Structure.Gray,System.Byte]
at ASToFra.scenProc.Steps.DetectFeatures.Process(List`1 cells, String[] arguments)
at ASToFra.scenProc.Processor.SceneryProcessor.Process()
at ASToFra.scenProc.Processor.SceneryProcessor.ProcessConfig(String filename, List`1 commands)

For Method 1, here is a screenshot:

The following steps were used for the above results:
ImportGDAL|B:\CA\Section_74\Images\CA_BS32_DEM.tif|*|NOREPROJ|String;CENTER|elev
ImportGDAL|B:\CA\Section_74\Images\CA_BS32.tif|*|NOREPROJ|String;type|image

ProcessElevationRaster|type="elev"|SLOPE|String;type|slope

SplitGrid|LOD13|*

DetectFeatures|type="image"|B:\CA\Section_74\_Control Data\CA_Section_74_TFE_Water.tf2|String;WatType|water|NONE

AddAttributeSampleRaster|type="water"|type="slope"|CENTER|slope

# Unload all features where the slope is above a certain threshold
UnloadFeatures|slope_0>1

MergeGrid

ExportOGR|WatType="water"AND FAREA>200|ESRI Shapefile|B:\CA\Section_74\Masks Water Tests\CA_BS32_WM_CENT_1_200.shp|water

It almost appears at the Unload step, anything greater than 1% is retained, not removed... the opposite? As you can see those slopes are way greater than 1%.

 

[arno]

In the step where you add the slope attribute you are using type="water", while elsewhere you use Wat type="water". So I don't think a slope attribute is added and therefore nothing is unloaded as well.

In the event log you should see messages how many features were processed by the step.

 

[Clutch Cargo]

Ok, made the correction. Still no change. Here is the script and a screenie of the processing where no features were removed.

ImportGDAL|B:\CA\Section_74\Images\CA_BS32_DEM.tif|*|NOREPROJ|String;CENTER|elev
ImportGDAL|B:\CA\Section_74\Images\CA_BS32.tif|*|NOREPROJ|String;type|image

ProcessElevationRaster|type="elev"|SLOPE|String;type|slope

SplitGrid|LOD13|*

DetectFeatures|type="image"|B:\CA\Section_74\_Control Data\CA_Section_74_TFE_Water.tf2|String;type|water|NONE

AddAttributeSampleRaster|type="water"|type="slope"|CENTER|slope

UnloadFeatures|slope_0>1

MergeGrid

ExportOGR|type="water"|ESRI Shapefile|B:\CA\Section_74\Masks Water Tests\CA_BS32_WM_CENT_1.shp|water

 

[arno]

You assign your elevation raster an attribute CENTER="elev", but when calculating the slope you select type="elev". So that means nothing matches and no slope is calculated.

 

[jpfil]

Hi,

Could the procedure be adapted for mountain roads in FS2020? Like lakes and rivers, we often see them following the slope when in reality they are transversely flat.

I had suggested this simple routine to Asobo: every x meters (depending on the desired degree of precision), transversely, place the highest side of the road at the same level as the lowest side. Without answer...

 

[arno]

I'm not sure how scenproc could help with roads in MSFS, since scenproc can not make them.

To be able to fix roads you would need very high resolution elevation data as well, which is not always available. It would be easier to have some procedural rules that roads are flat and modify the elevation data based on that.

 

[Clutch Cargo]

Back from another trip, I was able to make the adjustments and I am beginning to see promising results as a work-around. Lot's of testing in different geographic areas are still needed but this could save a lot a of "clean-up" time where the current water detection misses. I'll show you results when I continue testing.

 

[Clutch Cargo]

Starting to test with your 2nd method. I am confused with this step:

DetectFeatures|type="slope"|slope_threshold.tfc|String;type|mask|NONE

Am I to create a 2nd .tfc file (which you named above 'slope_threshold.tfc' as an example), for this method to process? If so, how does one create this tfc file? I take it using the TFE but is there a step to create this? Does one create a TFE script and use the Threshold Binary step? How would the TFE script look like?

 

[arno]

You just make a texture filter like normal, but then to work on the slope image instead of images. So it just needs a binary threshold step with the threshold slope value you want.

 

[Clutch Cargo]

I may need a little "hand-holding" as I go through method #2. To create the 2nd texture filter (.tf2), I will have to break down my DEM image into sample size images like I do with my regular images? Or can I just use the DEM image itself since the size is so small? Yea, way lost here. So if I open the Threshold Binary step in TFE I default threshold of 128, which would understand this is half way on the grayscale of the DEM image? How does the relate to slop when that represents altitude?

 

[arno]

Hi,

I would just try it at the current size first. You are not using the DEM image directly of course in the texture filter, you are using the slope image that was generated from it. By using the 2nd texture filter you can select the areas with low slope and use these for masking in the water detection.