Improving visual quality of synthetic terrains using weathering simulation on graphics processing unit

Visual simulation of natural eroded terrains using computer graphics has always been a fascinating research field. While many algorithms have been developed to improve terrain visual quality, physically-based hydraulic erosion algorithms have gained much attention in recent years for its ability to...

全面介紹

Saved in:
書目詳細資料
主要作者: Nguyen, Hoang Anh
其他作者: Parimal Aswani
格式: Theses and Dissertations
語言:English
出版: 2008
主題:
在線閱讀:https://hdl.handle.net/10356/14523
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
實物特徵
總結:Visual simulation of natural eroded terrains using computer graphics has always been a fascinating research field. While many algorithms have been developed to improve terrain visual quality, physically-based hydraulic erosion algorithms have gained much attention in recent years for its ability to generate realistic natural-looking terrains. In spite of that, most of the algorithms proposed so far have only been tested with low resolution terrains. When these algorithms are used on higher resolution terrains, they can take up to several hours to run, which make them unsuitable for many applications, especially real-time ones. That is the reason why in many graphics applications such as games nowadays, terrain is mostly generated offline and only loaded during game play. This situation will restrict the number of terrains to be stored when we have limited storage capacity. Recently, graphics hardware has become a new mechanism to improve the speed of computation. This has given us the idea to develop the next generation of General Pur-pose Computation on Graphics Processing Unit (GPGPU) erosion algorithms for faster terrain generation.In this thesis, we propose a fast and efficient GPU-based hydraulic erosion procedural technique which runs at interactive rates on today’s computers. This procedural technique can be used to abstract low resolution terrain data and generate high resolution realistic terrains on the fly during the application loading time.