基于HBase的矢量数据云存储策略研究  

Research on Cloud Storage of Vector Data Based on HBase

作　　者：祝若鑫 郭建忠[1] 薛帅[1] 范建永[1] 陈科[1] ZHU Ruoxin GUO Jianzhong XUE Shuai FAN Jianyong CHEN Ke(Information Engineering University, Zhengzhou 450001, China)

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《测绘科学技术学报》2016年第5期524-528,共5页Journal of Geomatics Science and Technology

基　　金：国家自然科学基金项目(41271392;41401462);郑州市科技攻关项目(112PPTGY225)

摘　　要：大数据时代,矢量数据量急剧增长,迫切需要寻找有效的矢量大数据存储方法。提出了一种基于HBase的矢量数据云存储策略。首先应用四叉树剖分方法构建多级格网索引并基于Hilbert填充曲线对矢量数据进行聚类划分;其次结合多级格网编码和Hilbert排列码设计矢量要素唯一标识并以此作为矢量要素在HBase数据库的行键;再次对矢量数据的存储规则进行了具体设计;最后通过两个对比实验对矢量要素唯一标识聚类效果及矢量数据查询效率进行了验证,实验结果证明了该方法的可行性与有效性。Nowadays we enter the big data era. The amount of vector data is growing explosively. There is an urgent need for efficient storage method of vector big data. A cloud storage strategy of vector data based on HBase was proposed in this paper. Firstly, quadtree decomposition method was applied to build multi-level grid index and Hilbert space filling curve was applied to partition vector data. Secondly, vector element unique identifier was designed based on multi-level grid code and Hilbert sequence code, which was treated as RowKey of vector element in HBase. Thirdly, the storage rule of vector data was designed in detail. Finally, two contrast experiments were used to verify good feasibility and high efficiency of this proposed method.

关 键 词：HBase数据库 云存储 矢量数据 Hilbert填充曲线 行键 

分 类 号：P208[天文地球—地图制图学与地理信息工程]