Reconstruction and Rendering of Time-Varying Natural Phenomena

Abstract

While computer performance increases and computer generated images getever more realistic, the need for modeling computer graphics content is becoming stronger. To achieve photo-realism, detailed scenes have to be modeled, often with a significant amount of manual labour. Interdisciplinary research combining the fields of Computer Graphics, Computer Vision, and Scientific Computing has led to the development of (semi-)automatic modeling tools freeing the user of labour-intensive modeling tasks.The modeling of animated content is especially challenging. Realistic motion is necessary to convince the audience of computer games, movies with mixed reality content, and augmented reality applications. The goal of this thesis is to investigate automated modeling techniques for time-varying natural phenomena. The results of the presented methods are animated, three-dimensional computer models of fire, smoke and fluid flows.