GeoJSON based conflation from a subset of OSM data - Githubissues

hotosm / osm-merge

Merge data into existing OSM data.

GNU Affero General Public License v3.0

3 stars 0 forks source link

GeoJSON based conflation from a subset of OSM data #8

Closed spwoodcock closed 2 months ago

spwoodcock commented 3 months ago

Is your feature request related to a problem? Please describe.

As discussed, we currently need the entire OSM postgres database to conflate against.
This is great to possibly integrate into an microservice that sits alongside raw-data-api (which has the updated OSM database).
But more realistically for tools like FMTM, we need to conflate against a small extract from OSM data.

Describe the solution you'd like

The ideal scenario would be GeoJSON based conflation.
Input: ODK data converted to GeoJSON, including all tag info as properties.
Input: data extract geojson from raw-data-api, including tags as properties.
Output: a single geojson of geometries, with conflated tags.
Where the geometry or tag conflation has uncertainty, it would be good to flag with a confidence level, so the user can manually review the uncertain entries.

Describe alternatives you've considered

Adding an OSM database alongside services like FMTM, but this would be significant data duplication.
A future microservice to create a conflation API alongside raw-data-api.

rsavoye commented 2 months ago

conflator.py has been refactored to support this using and file from ODK and the data extract for FMTM.