高性能集群式遥感数据并行处理系统及关键技术浅析  

A Brief Analysis of High Performance Computing Cluster System and Key Technologies for Remote Sensing Data

作　　者：潘彤 段云龙 谭盼华 PAN Tong;DUAN Yunlong;TAN Panhua(Tianjin Xinda Surveying and Mapping Technology Co.,Ltd.,Tianjin 300000,China;The 52th Research Institute of China Electronics Technology Group Corporation,Hangzhou 311100,China;Tianjin Institute of Surveying and Mapping,Tianjin 300381,China)

机构地区：[1]天津信达测绘科技有限公司,天津300000 [2]中国电子科技集团公司第五十二研究所,浙江杭州311100 [3]天津市测绘院有限公司,天津300381

出　　处：《测绘与空间地理信息》2022年第4期141-146,共6页Geomatics & Spatial Information Technology

摘　　要：高性能集群并行处理技术已成为海量遥感数据快速处理的重要方法。本文给出了一套精简的遥感数据高性能集群处理系统的结构,阐述了系统的并行处理机制,并从遥感数据并行处理策略、作业与资源管理系统、可视化流程定制、高性能计算编程模型等多个方面对集群式遥感数据处理中的关键技术进行了分析与说明。该研究成果可为遥感处理专业人员和高性能集群系统的设计和开发人员提供一定的参考作用。High performance computing cluster parallel processing technology has been the important fast method for massive remote sensing data.This paper presents a brief and simple structure of HPC cluster,and expounds its mechanism of parallel processing.The key technologies such as parallel processing strategy of remote sensing data,job and resource management and scheduling system,visual and customization data processing flow,HPC programming model are analysed and explained.There is a certain reference effect to people who are occupied by remote sensing processing or people who design and develop HPC cluster system for the research of this paper.

关 键 词：集群系统 高性能计算 并行处理 遥感数据 

分 类 号：P237[天文地球—摄影测量与遥感]