Procedural Generation of Dungeons

The use of procedural content generation (PCG) is not un-common amongst games, it appears in
many applications. PCG algorithms notably present the advantage of providing potentially infinite
and un-predictable content for the players, which can significantly add value to the game. However,
implementing a good PCG algorithm that generates convincing results is not a trivial task. PCG
revolves around the concept of randomness that is usually implemented by using pseudo-random
number generators. Different PCG methods can be used to create content for specific kinds of
environments. These methods can vary greatly depending on the type of environment desired. This
thesis focuses on generating content for dungeon-like or maze-like environments by identifying the
most suitable methods and algorithms. The methods used for generating maze involve finding a
random spanning tree of a graph, hence the Prim’s and Kruskal’s algorithms can be used. Further
on, we can obtain a dungeon-like environment from a maze by applying modifications to it such as
adding loops and rooms.