IT lexicon › Programming › Polygon mesh

Polygon mesh

Programming På svenska → Updated: 2026-10-24

The standard way to represent a 3D shape in games: a set of vertices joined into faces, almost always triangles.

A polygon mesh describes a surface with three lists — vertices (positions in space), edges and faces (facets). Triangles dominate because a GPU rasterizes exactly that; modelling tools often work in quads that get split on export. Each vertex also carries a normal (for lighting) and a UV coordinate (for texturing). Data structures such as half-edge and winged-edge make it cheap to walk the neighbourhood around a vertex.

The format has been the backbone of real-time graphics since the 1970s — the Utah teapot (Martin Newell, 1975) is the canonical example. The alternatives describe volume rather than surface: a voxel grid, an implicit surface, or a point cloud. A mesh says nothing about the interior, which is cheap but makes destructible geometry awkward. Detail is managed with mesh simplification and swaps between LOD tiers at distance.

← Back to the lexicon