A technique for turning a model of the world stored in the computer's memory into an image. Part of computer graphics. The key aspects of polygonal rendering are that it:

  • Employs a local illumination model; in other words, the light at a point only depends on the local surface characteristics and lights in the scene, and not on other objects in the scene. This differentiates it from, say radiosity algorithms.
  • It is an object space technique that maps from polygons directly to pixels on the screen, rather than asking whether what colour a pixel should be and then looking for what bits in the scene affect it. This differentiates it from ray tracing.
  • All objects in the scene are represented as polygons. This differentiates it from ray tracing.

Polygon rendering is particularly well-suited to real-time applications (like games). Systems like OpenGL and Direct3D are implementations and APIs for polygon rendering systems (ok, they both support NURBS but at the end of the day the NURBS are converted to lines and polygons anyway). Polygon rendering can be easily implemented in hardware, and hence all 3D accelerators are in fact designed to speed up polygon rendering. A 3D accelerator will not help you for ray tracing applications, but it will make Quake (which is built on a polygon renderer) run faster.