期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

基于osgEarth的大规模流场并行计算方法

osgEarth-based parallel algorithm for large scale flow field
下载PDF
导出
摘要 全球海洋流场具有数据规模大、计算复杂度高等特点,传统的可视化线性积分卷积(LIC)算法计算时耗高,难以满足大规模流场的实时交互可视化需求。本文基于OSG提出了大规模流场的LIC并行计算可视化方法,计算方法以osgEarth开源渲染引擎为基础,利用GPU的高性能计算技术,构建统一的设备计算框架,实现了LIC算法的并行化以达到加速效果。实验结果表明:该算法的加速比能够提高50倍,可满足全球海洋流场实时交互可视化处理的需求。 Global ocean current field has large data scale,which causes high computational complexity when visualizing it.Because of its high computational time consumption,it is difficult for traditional linear integral convolution(LIC)algorithm to meet the real-time interactive visualization requirements of large-scale flow field.This paper proposes a parallel LIC visualization method based on OpenSceneGraph(OSG),an open source rendering engine.GPU's high-performance computing technology is used to construct a unified device computing framework to parallelize the LIC algorithm.Experimental results show that:compared with traditional method,the acceleration ratio of this study can reach 50 times,which can meet the real-time interactive visualization of the global ocean current.
作者 王云鹏 苏学娟 秦勃 孙苗 WANG Yun-peng;SU Xue-juan;QIN Bo;SUN Miao(Department of Computer Science&Technology,Ocean University of China,Qingdao 266100,China;National Marine Data and Information Service,Tianjin 300171,China;Key Laboratory of Digital Ocean,State Oceanic Administration,Tianjin 300171,China)
机构地区 中国海洋大学计算机科学与技术系 国家海洋信息中心 国家海洋局数字海洋科学技术重点实验室
出处 《海洋信息》 2018年第4期15-20,共6页 Marine Information
关键词 OSG osgEarth GPU PARALLEL COMPUTATION LIC CUDA OSG osgEarth GPU parallel computation LIC CUDA
分类号 TP391.9 [自动化与计算机技术—计算机应用技术]
  • 相关文献

参考文献1

二级参考文献14

  • 1俞宏峰.大规模科学可视化[OL]. [2014-03-06]. http://www.ccf.org.cn/sites/cc/zlcontnry.jsp-contentld=2694962949049.
  • 2Ueng S K, Sikorski C, Ma K L. Out-of-core streamline visualizationon large unstructured meshes[J]. IEEE Transactions onVisualization and Computer Graphics, 1997, 3(4): 370-380.
  • 3Kendall W, Huang J, Peterka T, Toward a general I/Olayer for parallel visualization applications[J]. IEEE Transactionson sualization and Conqjuter Graphics and Applications,2011,31(6): 6-10.
  • 4Chen Y, Byna S, Sun X H, et al. Hiding I/O latency with preexecutionprefetching for parallel applications [C] //Proceedingsof the ACM/IEEE Conference on Supercomputing. Los Alamitos:IEEE Computer Society Press, 2008: Article No. 40.
  • 5Guo H, Zhang J, Liu R, et al. Advection-based sparse datamanagement for visualizing unsteady flow[J]. IEEE Transactionson Msualization and Computer Graphics, 2014, 20(12):2555-2564.
  • 6Bruckschen R, Kuester F, Hamann B, et al. Real-time out-ofcorevisualization of particle traces[C] //Proceedings of theIEEE Symposium on Parallel and Large-data ^sualization andGraphics. Los Alamitos: IEEE Computer Society Press, 2001:45-50.
  • 7Nouanesengsy B, Lee T Y, Shen H W. Load-balanced parallelstreamline generation on large scale vector fields[J]. IEEETransactions on Visualization and Computer Graphics, 2011,17(12): 1785-1794.
  • 8Yu H F, Wang C L, Ma K L. Parallel hierarchical visualizationof large time-varying 3d vector fields[C] //Proceedings of theACM/IEEE Conference on Supercomputing. Los Alamitos:IEEE Computer Society Press, 2007: Article No. 24.
  • 9Chen L, Fujishiro I. Optimizing parallel performance ofstreamline visualization for large distributed flow datasets[C]//Proceedings of the Pacific Visualization Symposium. LosAlamitos: IEEE Computer Society Press, 2008: 87-94.
  • 10Pugmire D, Childs H, Garth C, et al. Scalable computation ofstreamlines on very large datasets[C] //Proceedings of the IEEEConference on High Performance Computing Networking,Storage and Analysis. Los Alamitos: IEEE Computer SocietyPress, 2009: Article No. 16.

共引文献4

海洋信息
2018年 第4期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部