首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种基于GPU Tessellation的地形无缝绘制算法
引用本文:董路明,张斌,赵学胜.一种基于GPU Tessellation的地形无缝绘制算法[J].武汉大学学报(信息科学版),2017,42(3):402-407.
作者姓名:董路明  张斌  赵学胜
作者单位:中国矿业大学 (北京) 地球科学与测绘工程学院, 北京, 100083
基金项目:国家自然科学基金41171306国家自然科学基金41171304
摘    要:针对传统多分辨率地形绘制算法构网速度慢、T型裂缝不易处理等问题,提出了一种GPU(graphic processing unit)构网的地形无缝绘制算法。首先,引入实时网格细分(tessellation)技术,将地形构网分为CPU粗粒度Tile网格构建和GPU细粒度Patch网格细分两个阶段;然后,Tile网格采用基于视距的细节层次模型进行LoD层次选择,Patch采用基于屏幕空间投影误差的细节层次模型完成网格细分,兼顾了视距和地形粗糙度对地形绘制的影响,实现了地形细节层次的自适应选择;最后,应用C++语言和DirectX 11工具,设计开发了相应的可视化实验系统。实验结果表明,该方法实现了多分辨率地形的自适应无缝表达,保证了地形网格的连续性;并通过合理平衡CPU-GPU负担,显著提升了地形渲染效率。

关 键 词:地形渲染    GPU    曲面细分    裂缝消除
收稿时间:2015-03-08

A Seamless Terrain Rendering Algorithm Based on GPU Tessellation
Institution:College of Geoscience & Surveying Engineering, China University of Mining & Technology (Beijing), Beijing 100083, China
Abstract:In this paper, a seamless terrain rendering algorithm based on GPU is proposed to solve the problem that the conventional multi-resolution terrain rendering methods are time-consuming and not suitable for processing T-crack. Firstly, the real-time Tessellation is introduced and the terrain mesh constructing is divided into two steps, i.e. conducting coarse granularity Tile mesh on CPU and subdividing fine granularity Patch mesh on GPU. Then, in order to take the range of visibility and the terrain roughness on terrain rendering into account simultaneously, in each Tile level of detail is selected according to LoD model based on the range of visibility and in each Patch level of subdivision is selected according to LoD model based on screen-space projection error. Thus, in this algorithm the selection of the terrain LoD is realized self-adaptively. Finally, by using C+ + language and DirectX 11 tools, a visual experiment system is developed and the experimental result shows that the multi-resolution terrain can be expressed seamlessly and self-adaptively ensuring the continuity of the terrain mesh. Moreover, our algorithm can significantly increase the terrain rendering efficiency through reasonable balance of CPU and GPU.
Keywords:
本文献已被 CNKI 等数据库收录!
点击此处可从《武汉大学学报(信息科学版)》浏览原始摘要信息
点击此处可从《武汉大学学报(信息科学版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号