English

A Coreset for Approximate Furthest-Neighbor Queries in a Simple Polygon

Computational Geometry 2024-03-08 v1

Abstract

Let P\mathcal{P} be a simple polygon with mm vertices and let PP be a set of nn points inside P\mathcal{P}. We prove that there exists, for any ε>0\varepsilon>0, a set CP\mathcal{C} \subset P of size O(1/ε2)O(1/\varepsilon^2) such that the following holds: for any query point qq inside the polygon P\mathcal{P}, the geodesic distance from qq to its furthest neighbor in C\mathcal{C} is at least 1ε1-\varepsilon times the geodesic distance to its further neighbor in PP. Thus the set C\mathcal{C} can be used for answering ε\varepsilon-approximate furthest-neighbor queries with a data structure whose storage requirement is independent of the size of PP. The coreset can be constructed in O(1ε(nlog(1/ε)+(n+m)log(n+m)))O\left(\frac{1}{\varepsilon} \left( n\log(1/\varepsilon) + (n+m)\log(n+m)\right) \right) time.

Keywords

Cite

@article{arxiv.2403.04513,
  title  = {A Coreset for Approximate Furthest-Neighbor Queries in a Simple Polygon},
  author = {Mark de Berg and Leonidas Theocharous},
  journal= {arXiv preprint arXiv:2403.04513},
  year   = {2024}
}

Comments

To appear in SoCG 2024

R2 v1 2026-06-28T15:12:21.403Z