Interactive Procedural Terrain Generation System

Abstract

This thesis presents the design and implementation of an interactive procedural terrain generation system that aims to balance visual realism, computational efficiency, and user control. The developed system uses noise-based terrain generation methods, including Perlin noise, Fractal Brownian Motion, and Ridged Multi-Fractal techniques, to generate diverse landscapes. To improve realism, a Lagrangian hydraulic erosion simulator was integrated to create landforms that are caused by the effect of water flow on natural terrain. A texturing component is presented which applies either UV mapping or triplanar box mapping to apply textures to the terrain. Multiple strategies are introduced to blend different textures on the terrain surface based on height and slope to realistically apply materials such as grass, rock, and snow. Performance was further improved through an optimized normal vector computation method, which achieved significantly faster execution than traditional approaches while maintaining high accuracy. The final system demonstrates that combining procedural algorithms with physically inspired simulation can provide a flexible and effective solution for real-time terrain creation in graphics applications.