基于数据分解的GPU加速地形分析并行计算  

GPU accelerated parallel computing for terrain analysis based on data decomposition

作　　者：高凯 李雯[4] 郭菲 GAO Kai;LI Wen;GUO Fei(School of Geospatial Information,Information Engineering University,Zhengzhou 450001,China;Xi'an Research Institute of Surveying and Mapping,Xi'an 710054,China;State Key Laboratory of Geo-Information Engineering,Xi'an 710054,China;Xi'an Aerospace Remote Sensing Data Technology Co.,Ltd.,Xi'an 710061,China)

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450001 [2]西安测绘研究所,陕西西安710054 [3]地理信息工程国家重点实验室,陕西西安710054 [4]西安航天天绘数据技术有限公司,陕西西安710061

出　　处：《测绘科学与工程》2019年第6期32-35,共4页Geomatics Science and Engineering

摘　　要：数字地形分析中越来越庞大的数据量和越来越复杂的计算模型,对现有传统的串行计算模式提出了挑战。为了提高地形分析模型计算效率和应对海量数据,本文提出了基于GPU硬件加速的地形分析并行计算模式,并研究了使GPU利用率达到最佳的配置方法,通过实验验证了基于GPU硬件加速的地形分析并行计算的可行性和高效性。The increasingly large amount of data and complex computing models in digital terrain analysis have posed a challenge to the traditional serial computing model.To improve the computing efficiency of terrain performance analysis model and better process massive data,a GPU hardware accelerated parallel computing model for terrain analysis is proposed in this paper.Meanwhile,the best configuration method for GPU utilization is studied.Experimental results show that the GPU hardware accelerated parallel computing for terrain analysis is feasible and efficient.

关 键 词：地形分析 GPU 并行计算 数据分解 DEM 通视分析 

分 类 号：P208[天文地球—地图制图学与地理信息工程]