基于Geo SOT的分布式遥感图像高效存储与检索  

Efficient Storage and Retrieval of Distributed Remote Sensing Images Based on GeoSOT

作　　者：刘钢 马俊 徐玮[2] 刘煜 LIU Gang;MA Jun;XU Wei;LIU Yu(School of Information Science and Engineering,Hunan Institute of Science and Technology,Yueyang 414006,China;College of Systems Engineering,National University of Defense Technology,Changsha 410073,China)

机构地区：[1]湖南理工学院信息科学与工程学院,湖南岳阳414006 [2]国防科技大学系统工程学院,湖南长沙410073

出　　处：《湖南理工学院学报(自然科学版)》2025年第1期14-17,共4页Journal of Hunan Institute of Science and Technology(Natural Sciences)

基　　金：湖南省自然科学基金项目(2024JJ5173,2023JJ50047);湖南省教育厅科学研究重点项目(23A0494)。

摘　　要：随着遥感技术的快速发展,遥感图像数据量呈指数级增长.为解决海量遥感图像数据的存储和快速索引问题,设计一种基于Hadoop集群的分布式遥感图像存储架构,搭建全球位置栅格(Geo SOT)轻量级金字塔模型,并使用Geo SOT框架实现基于Z阶曲线的分布式索引技术.预处理的遥感图像最初存储在Hadoop分布式文件系统中,然后使用GeoTrellis分布式计算引擎构建金字塔模型,并将生成的结果保存在分布式数据库Accumulo中,为每个瓦片数据添加基于Geo SOT实现的Z阶曲线编码空间索引.实验表明,这种存储体系结构能有效提高遥感图像的访问和读取效率.与Windows上运行的Arc GIS和Postgresql以及原生GeoTrellis相比,金字塔模型创建和数据库集成所需的时间显著减少.此外,其读取效率大大超过传统的空间数据库.With the rapid development of remote sensing technology,the volume of remote sensing image data has experienced exponential growth.In order to address the challenges of storing and quickly indexing massive remote sensing images,a distributed remote sensing image storage architecture based on the Hadoop cluster has been designed.This architecture incorporates a lightweight pyramid model called the Global Position Grid(Geographical coordinates Subdividing grid with One dimension integral coding on 2n-Tree,GeoSOT).The GeoSOT framework is utilized to implement a distributed indexing technique based on the Z-order curve.Preprocessed remote sensing images are initially stored in the Hadoop Distributed File System(HDFS).Subsequently,the GeoTrellis distributed computing engine is employed to construct the pyramid model,and the generated results are stored in the distributed database Accumulo.A spatial index based on Z-order curve encoding,implemented through GeoSOT,is added to each tile’s data.The experimental validation demonstrates that this storage architecture effectively enhances the access and retrieval efficiency of remote sensing images.The time required for pyramid model creation and database integration is significantly reduced compared to running on Windows with ArcGIS and Postgresql,as well as native GeoTrellis.Furthermore,the reading efficiency surpasses traditional spatial databases by a considerable margin.

关 键 词：GeoSOT Z阶曲线 分布式存储 轻量级金字塔模型 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] P237[自动化与计算机技术—控制科学与工程]