Graphics in Flatland: a case study

Usually in computer graphics, a two-dimensional view of a set of three-dimensional objects is considered. In this article we reduce the dimensionality by one in each case. In other words we study what, for obvious reasons, we call Flatland graphics. This forms the beginning of a mathematical investigation of computer graphics and, at the same time, provides uniform solutions for a number of computational geometry problems. In particular we study the maintenance of a view during insertion and deletion of objects and the "frame-to-frame" coherence while walking around a set of objects. Both parallel and perspective projections are considered. Our major concern is convex objects that are simple—in a sense, made precise in this article. However, we will close this article by discussing some possible extensions to nonconvex objects and/or to higher dimensions. The investigation also serves to demonstrate a number of tools that have been developed recently in the context of computational geometry. For example. dynamization and searching.