Delaunay Triangulation Simulator: Optimal Mesh Generation

simulator intermediate ~10 min
Loading simulation...
46 triangles from 25 points

25 points produce approximately 46 Delaunay triangles with guaranteed maximized minimum angle — the highest-quality triangulation possible for this point set.

Formula

F = 2n - 2 - b — triangle count (b = boundary vertices)
E = 3n - 3 - b — edge count
circumcircle: no point pk inside circle(pi, pj, pl)

The Optimal Triangulation

Given a set of points, there are exponentially many ways to triangulate them. The Delaunay triangulation is special: it maximizes the minimum angle across all triangles, producing the 'fattest' possible mesh. This property, formalized by Boris Delaunay in 1934, makes it the gold standard for mesh generation in engineering simulation, terrain modeling, and computer graphics.

The Empty Circumcircle Property

Every Delaunay triangle has a circumscribed circle containing no other points. This elegant geometric invariant serves as both definition and algorithm: when inserting a new point violates the property, 'flip' the offending edge to restore it. The simulator highlights circumcircles in real time, letting you verify the property visually as points move.

Algorithms and Complexity

Three main algorithms construct Delaunay triangulations: incremental insertion (add points one by one, flipping edges), divide and conquer (split, triangulate halves, merge), and Fortune's sweep line (adapted from Voronoi computation). All achieve O(n log n) time — provably optimal since Delaunay triangulation can sort numbers.

From Meshes to Mountains

Delaunay triangulation powers countless applications. Finite element solvers need well-shaped meshes to avoid numerical instability — Delaunay guarantees this. Terrain models (TINs) use Delaunay triangulation of elevation points to create smooth, artifact-free surfaces. And in 3D reconstruction, Delaunay tetrahedralization extends the concept to volumetric meshing for medical imaging and structural analysis.

FAQ

What is a Delaunay triangulation?

A Delaunay triangulation of a point set is a triangulation where no point lies inside the circumscribed circle of any triangle. This uniquely maximizes the minimum angle across all triangles, producing 'fat' well-shaped triangles ideal for numerical computation. Named after Boris Delaunay (1934).

What is the empty circumcircle property?

For every triangle in a Delaunay triangulation, its circumscribed circle contains no other points from the set. This property is both the definition and the key algorithmic invariant — when violated, an edge flip restores it.

How is Delaunay triangulation related to Voronoi diagrams?

They are duals: connect two Voronoi seeds whenever their cells share an edge, and you get the Delaunay triangulation. Conversely, the perpendicular bisectors of Delaunay edges form the Voronoi diagram. Computing one gives you the other for free.

Why is Delaunay triangulation important?

It produces optimal-quality meshes for finite element analysis, terrain modeling, 3D surface reconstruction, and pathfinding. The maximum-minimum-angle property prevents the numerical errors that arise from thin, elongated triangles in computational simulations.

Sources

Embed

<iframe src="https://homo-deus.com/lab/computational-geometry/delaunay-triangulation/embed" width="100%" height="400" frameborder="0"></iframe>
View source on GitHub