You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We calculate its concave shell with ST_ConcaveHull(geom, param_pctconvex, param_allow_holes):
As described in the document, the param_pctconvex controls the concaveness of the computed hull. A value of 0 produces a hull with maximum concaveness (but still a single polygon). And often values between 0.3 and 0.1 produce reasonable results.
nbvfgh
changed the title
ST_ConcaveHull gives unexpected result with MultiPolygon
ST_ConcaveHull gives unexpected results of MultiPolygon
Dec 23, 2024
dr-jts
changed the title
ST_ConcaveHull gives unexpected results of MultiPolygon
ST_ConcaveHull of MultiPolygon returns disconnected result
Dec 23, 2024
This is a limitation of the current algorithm. It can produce disconnected results for low threshold values.
It is possible to fix this, at the cost of some performance. It requires checking that the internal triangulation remains connected after each triangle "erosion".
Considering following MultiPolygon:
We calculate its concave shell with ST_ConcaveHull(geom, param_pctconvex, param_allow_holes):
As described in the document, the param_pctconvex controls the concaveness of the computed hull. A value of 0 produces a hull with maximum concaveness (but still a single polygon). And often values between 0.3 and 0.1 produce reasonable results.
Case param_pctconvex = 0.47:
Case param_pctconvex = 0.25:
Case param_pctconvex = 0:
The above three examples do not meet expectations. When param_pctconvex is set to 0, it even degenerates into three original polygons
Version Info:
GEOS="3.13.0-CAPI-1.19.0
The text was updated successfully, but these errors were encountered: