On 10/30/22 10:15, Mike Thompson wrote:
One small thing on the proposal, it does seem like you have filtered out
some buildings from ms-kiowa.geojson that do not exist in
osm-Kiowa.geojson (in other words, they do not appear in
Kiowa-result.geojson). For example the MS building at this
location 38.481191, -102.778559
I've been going through the data deeper this morning, and validating
with imagery. Many grain silos are mi-identified as buildings, and the
usual bug in Bing footprints, all converted to squares. :-( Luckily
they're often in groups, and usually near farms with lots of other
buildings, but I'll have to be careful. And there's also some great
Salvador Dali "footprints" in farmers fields. Changes in topography are
also getting misidentified, but easy to find. I'm getting some
interesting screenshots for a future conference presentation. :-)
Nice catch ! I'm not sure why that building got skipped, I'll have to
check my software. There could be others. I may import what is in the
current results since I already started cleaning it up, then make
another pass as a followup to get anything missed. I'll make sure the
bug is fixed before doing any other counties. I've been mostly focused
on making sure I wouldn't be adding any duplicates.
This part of Colorado is definitely a desert of OSM data...
- rob -
--
https://www.senecass.com
_______________________________________________
Imports mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/imports
