四近邻驱动的受限Voronoi图快速构造方法
Fast Construction Method for Restricted Voronoi Diagrams Driven by Four Nearest Neighbors
-
摘要: 受限Voronoi图广泛应用于计算机图形学等领域, 提升其构造效率对下游应用具有重要意义. 针对现有的主流构造方法存在部分裁剪对结果没有贡献的问题, 提出一种四近邻驱动的受限Voronoi图快速构造方法. 通过扩散方式迭代地搜索Voronoi胞元与网格面片对, 并依次计算每对的交集. 所提方法的创新点在于Voronoi胞元与网格面片之间交集的高效计算. 遍历面片的每条边, 若两个端点所属的Voronoi胞元不同, 则通过几次四近邻搜索确定目标种子点与其邻居的中垂面, 并对面片进行裁剪. 与主流方法的对比实验结果表明, 所提方法具有更高的效率, 且对种子点分布的敏感程度显著降低. 在种子点和网格面片数量均为50k且种子点分布随机的情况下, 所提方法相比两种主流方法分别减少了71%和30%的耗时.Abstract: Restricted Voronoi diagrams are widely used in fields such as computer graphics, and improving their construction efficiency is important for downstream applications. Addressing the issue in mainstream methods where partial clippings contribute nothing to the final result, this paper proposes a fast construction method driven by four-nearest-neighbor searches. It iteratively searches for Voronoi cell-mesh face pairs via propagation and sequentially computes their intersections. The core innovation lies in the efficient computation of intersections between Voronoi cells and mesh faces. For each edge of a face, if its two endpoints belong to different Voronoi cells, a bisector between the target site and one of its neighbors is found through several four-nearest-neighbor searches and used to clip the face. Comparative experiments with mainstream methods demonstrate that the proposed method achieves higher efficiency and exhibits significantly reduced sensitivity to site distribution. In tests with 50k sites and 50k mesh faces under random site distribution, the proposed method reduced computation time by 71% and 30% respectively compared to the two mainstream approaches.