Point in Polygon quadtree version

[Schubert-Tom]

Hi there,

I am working on a project where I want to do efficient point in polygon checks. I have one big polygon which I slice down in multiple small polygons at the beginning of my program.

After that the set of polygons stays constant but I get new random set of points all the time and have to check as fast as possible which of those points are within which polygon of the set.

For starting I just used the normal point_in_polygon function. I saw that there is some spatial indexing with quadtrees available but as of my understanding the quadtree is build for a set of points in a bounding box. After that this quadtree can be used with the quadtree version of the point in polygons function to check the position of those points relative to some polygons.

I was wondering if there is a possibility to build spatial indexing for the set auf polygons which stays constant and therefore speed up the inside polygonscheck by simply keeping the indexed tree of the Polygons in gpu memory?

Is there support for such a use case?
Can somebody recommend an approach for such a problem?

Thanks a lot in advance!

P.S
Each polygon (10 in total have ~70 points and the set of random points has around 250 points