Procedural generation and games: Implementation and performance of a survival video game based on a procedurally generated map

Abstract

A spectrum of games implemented by procedural generation takes one of the most influential positions in the whole industry. They always have an element of surprise behind their back and are willing to entertain a player with a different outcome after each fresh start, unlike games with the already prepared scenario. Nevertheless, the amount of information introduced to get acknowledged with procedural generation is extremely blurred in the context of a variety of solutions and algorithms, so it becomes puzzling to choose the right approach. What is more, this causes the knowledge to become more and more inaccessible in the context of the difficulty of the topic. That’s why this thesis describes the research of creating a map for a survival game by using the Perlin Noise algorithm as a foundation of the aforementioned technology. The requirements for this project are based on the scrupulous acknowledgment of aspects of coherent noise, its behavior, and adjustments that we can apply to achieve desirable height maps and assign necessary terrains accordingly. Based on the investigations above the theory was implemented into a working algorithm by using C# language together with Unity Game Engine. As a result of the implementation, the algorithm allowed adjusting the size of the generated map, as well as it was flexible towards the number of possible biomes that could appear during the fresh start of the game. In addition, the project made it possible to work with the random level generator of game objects and overall optimization management. However, the achievement of biome-sensitive objects wasn’t successful due to the nature of the Perlin Noise algorithm.