
Am 09.01.2010 00:05, schrieb Johann Gail:
Changes: * Polygons with inner polygons are now divided into several polygons. It's much faster and the PolygonSplitter cannot destroy them any more.
Hello WanMil,
this change could have some other side effects. (As observed and described in another thread) If the polygons gets small enough, then the SizeFilter will remove parts of them. I.e. if a big forrest gets cut into smaller pieces, some of them may be small enough to be catched by the SizeFilter.
Is it a must for splitting them up? Or was it just done to serve the PolygonSplitter?
Regards, Johann
Johann, the old findCpa method was quite slow. Its complexity is O(n^3) at its worsty worst case and usually O(n^2). Therefore I started to implement the new algorithm without the findCpa method. The current implementation may have problems with small polygons but it should be possible to perform some remerging. Another idea is to add some specific tags by the mulitpolygon algorithm that link to the other pieces of the formerly big forrest. This could be evaluated by the SizeFilter? WanMil