Thanks peter! It looks like this ground has been blazed by fine minds already!
Takeaways from that thread:
- R-trees can defeat the the O(n^2) bound. (Rhino.Geometry.RTree, wiki, paper, Radovan’s code, Qythium’s code)
- Impala provides components for faster curve-curve intersection and point-curve containment (and it’s open source!)
- Offsets/Minkowski sums could help cluster the inside curves