Pathfinding in Procedurally Generated Mazes
Matej Kuprešak, Časlav Livada, Tomislav Galba, Alfonzo Baumgartner · 2024
This paper presents the development of a simulator for procedural generation of mazes and evaluation of pathfinding algorithms. The implementation includes a Perlin noise-based procedural generation where the generated noise determines the colors of the terrain, increasing the realism of the terrain. The Unity interface allows the user to control and change these properties. Two pathfinding algorithms, namely Dijkstra’s algorithm and the breadth-first search algorithm, are used to test and analyze their effectiveness. The algorithms use positional data to identify and display feasible paths between specific points. The integrated user interface provides a straightforward platform to experiment with the algorithms and visualize their results, including the number of nodes in the path and the duration of pathfinding. In addition, a graphical representation allows real-time tracking of the player’s movement along the selected path. The results presented provide valuable insights for selecting algorithms for specific scenarios. Overall, this simulator is a valuable tool for researchers and developers involved in maze generation and pathfinding algorithm evaluation.