A Time-Optimal Delaunay Refinement Algorithm in Two Dimensions

Sariel Har-Peled,Alper Ungor

DOI: https://doi.org/10.48550/arXiv.cs/0501007

2005-01-05

Abstract:We propose a new refinement algorithm to generate size-optimal quality-guaranteed Delaunay triangulations in the plane. The algorithm takes $O(n \log n + m)$ time, where $n$ is the input size and $m$ is the output size. This is the first time-optimal Delaunay refinement algorithm.

Computational Geometry

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to propose a new optimization algorithm to generate Delaunay triangulations in the plane with optimal size and quality guarantees. Specifically, the author aims to develop a time - optimal Delaunay refinement algorithm with a running time of \(O(n \log n + m)\), where \(n\) is the input size and \(m\) is the output size.

### Background and Problem Description of the Paper

In the discretization of the geometric domain (i.e., meshing), avoiding small and large angles is crucial for reducing numerical and interpolation errors. Delaunay triangulation becomes a powerful discretization tool by maximizing the minimum angle. However, depending on the input configuration, Delaunay triangulation may have arbitrarily small angles. Therefore, researchers have developed Delaunay refinement algorithms with iterative insertion of additional points to solve this problem.

### Existing Problems and Improvements

Although there are already various Delaunay refinement algorithms, their time complexity is relatively high. For example, the original Delaunay refinement algorithm has a quadratic time complexity [Rup93]. In contrast, the quadtree refinement algorithm [BEG94] has a time complexity of \(O(n \log n + m)\) and is more efficient. To improve efficiency, Spielman et al. [STÜ02] proposed a parallelized Delaunay refinement algorithm, and Miller [Mil04] further improved the time complexity of the sequential algorithm.

### Contributions of This Paper

This paper presents the first time - optimal Delaunay refinement algorithm. This algorithm uses so - called "off - centers" instead of traditional circumcenters. The off - centers are located on the angular bisectors of the shorter sides and are closer to the shortest side. Experiments show that this method can significantly reduce the number of inserted Steiner points and generate smaller meshes. In addition, the new algorithm avoids the high - cost operation of maintaining the entire Delaunay triangulation and instead uses a scaffold quadtree structure to efficiently calculate, locate, and insert off - centers.

### Main Conclusions

1. **Time Complexity**: The time complexity of the new algorithm is \(O(n \log n + m)\), which is the same as that of the quadtree refinement algorithm.
2. **Quality Guarantee**: The triangulation generated by the new algorithm has a quality guarantee, and the generated mesh is smaller compared to existing methods.
3. **Locality**: The off - center insertion is a more local operation, reducing the impact on the global structure.

In conclusion, this paper solves the optimization problems of Delaunay refinement algorithms in terms of time and quality and provides an efficient and high - quality mesh generation method.