I have many circles which are lat long positions and a radius in meters.
I need to check if two circles intersect and I can use this haversine distance function.
However, to speed up performance I would like to build a quadtree containing all the circles.
From what I understand, I need to project the lat long positions into a projection that doesn't have distance distortion.
This would allow me to a) check circle intersection using a normal dist function, but most importantly b) allow me to check circle and rectangle intersection to build up a quadtree.
I've been searching for ages and couldn't find any way of doing such projection?
I understand that such projection would result in an oval space (I think) instead of a rectangular space, and therefore the quadtree implementation wouldn't be perfect, but enough to improve performance.
Also, if you know any other way to build quadtree of such circles, please let me know.
This question looked promising but couldn't find any answer: https://stackoverflow.com/questions/23780072/quadtree-for-collisions-with-latitude-longitude-earth-size