Dissolve Overlapping Polygons (with GDAL/OGR) while keeping non-connected results distinct

Question

Is there any way to dissolve (merge) overlapping polygons, using any GDAL/OGR API or command line tool, while keeping the resulting non overlapping areas distinct? I have searched a lot and I cannot find anything resembling what need. However, I think it is very unlikely that this problem has not been solved yet.

Here's a more detailed description of what I need:

• My input consists of a single shape file (ESRI Shapefile) with a single layer.
• This layer contains polygons which are not distinguishable by attributes. (All have the same attributes).
• Many of them are overlapping and I would like to get the union of those who are overlapping.
• Areas that are not connected should result in separate polygons.

It is the last point which is causing troubles. I basically get what I need except for the last point. If I run the typical solution for dissolving a shape file

$ ogr2ogr -f "ESRI Shapefile" dissolved.shp input.shp -dialect sqlite -sql "select ST_union(Geometry) from input"

I end up with a single polygon which includes everything, even if the areas are not connected.

Update: I solved the problem by ditching GDAL entirely. As many sources point out it is generally a better approach to use fiona and shapely to work with shapefiles. I have posted my solution below.

Answer 1

I had a similar problem and solved it like this, taking all the previous answers from Taking Union of several geometries in GEOS Python? into account:

import os
from osgeo import ogr

def createDS(ds_name, ds_format, geom_type, srs, overwrite=False):
    drv = ogr.GetDriverByName(ds_format)
    if os.path.exists(ds_name) and overwrite is True:
        deleteDS(ds_name)
    ds = drv.CreateDataSource(ds_name)
    lyr_name = os.path.splitext(os.path.basename(ds_name))[0]
    lyr = ds.CreateLayer(lyr_name, srs, geom_type)
    return ds, lyr

def dissolve(input, output, multipoly=False, overwrite=False):
    ds = ogr.Open(input)
    lyr = ds.GetLayer()
    out_ds, out_lyr = createDS(output, ds.GetDriver().GetName(), lyr.GetGeomType(), lyr.GetSpatialRef(), overwrite)
    defn = out_lyr.GetLayerDefn()
    multi = ogr.Geometry(ogr.wkbMultiPolygon)
    for feat in lyr:
        if feat.geometry():
            feat.geometry().CloseRings() # this copies the first point to the end
            wkt = feat.geometry().ExportToWkt()
            multi.AddGeometryDirectly(ogr.CreateGeometryFromWkt(wkt))
    union = multi.UnionCascaded()
    if multipoly is False:
        for geom in union:
            poly = ogr.CreateGeometryFromWkb(geom.ExportToWkb())
            feat = ogr.Feature(defn)
            feat.SetGeometry(poly)
            out_lyr.CreateFeature(feat)
    else:
        out_feat = ogr.Feature(defn)
        out_feat.SetGeometry(union)
        out_lyr.CreateFeature(out_feat)
        out_ds.Destroy()
    ds.Destroy()
    return True

UnionCascaded requires MultiPolygon as a geometry type, which is why I implemented the option to re-create single polyons. You could also use ogr2ogr from the command line with option -explodecollections:

ogr2ogr -f "ESRI Shapefile" -explodecollections dissolved.shp input.shp -dialect sqlite -sql "select ST_union(Geometry) from input"

Answer 2

So, after many unsuccessful attempts I have ditched gdal/ogr and went on with shapely and fiona. This does exactly what I need. The filtering was necessary becuase my dataset contains self-intersecting polygons which need to be filtered out before calling cascaded_union.

import fiona                                                                                                       
from shapely.ops import cascaded_union                                                                             
from shapely.geometry import shape, mapping  

with fiona.open(src, 'r') as ds_in:                                                                                                                                                                                                                   
    crs = ds_in.crs 
    drv = ds_in.driver 

    filtered = filter(lambda x: shape(x["geometry"]).is_valid, list(ds_in))                                                                                   

    geoms = [shape(x["geometry"]) for x in filtered]                                                   
    dissolved = cascaded_union(geoms)                                    

schema = {                                                                                                     
    "geometry": "Polygon",                                                                                     
    "properties": {"id": "int"}                                                                                
}  

with fiona.open(dst, 'w', driver=drv, schema=schema, crs=crs) as ds_dst:                                       
    for i,g in enumerate(dissolved):                                                                           
        ds_dst.write({"geometry": mapping(g), "properties": {"id": i}})