Quality indicators of LPIS processor

Dear Sen4CAP team,

I would like to know what would happen to a parcel

  1. if it is duplicated with the other parcel. Would it be still used and calculated one more time in the L4X processor?

  2. If two parcels overlap with each other partly, would the L4X processor still calculate the results for both parcels?

Best regards,
Shengyin

Dear Shengyin,

The L4A processor excludes duplicate and overlapping parcels (and also those with invalid geometry), so they are never classified (not even once). The L4C (and L4B, I think) processors ignore those flags and process all the duplicates.

Hallo Inicola,

Thank you very much for your answer!
Is it possible to change any configuration, and let L4A also ignore those flags, and process the overlapping parcels as well?

Best regards,
Shengyin

Not right now – you’d have to patch a Python script for that:

diff --git i/scripts/extract-parcels.py w/scripts/extract-parcels.py
index 07276f8b..c76cd527 100755
--- i/usr/bin/extract-parcels.py
+++ w/usr/bin/extract-parcels.py
@@ -78,7 +78,6 @@ def extract_parcels(config, args, lpis_table, lut_table, id, geom):
             natural join {} lut
             where "GeomValid"
             and not "Duplic"
-            and not "Overlap"
             and not is_deleted
             order by "NewID"
             """
@@ -98,7 +97,6 @@ def extract_parcels(config, args, lpis_table, lut_table, id, geom):
             natural join {} lut
             where "GeomValid"
             and not "Duplic"
-            and not "Overlap"
             and not is_deleted
             order by "NewID"
             """

But this generally doesn’t work very well because of technical reasons. It’s very likely that the overlapping parcels (with the exception of an arbitrary one) will not be classified.