空间索引技术及其GIS应用综述

A Review of Spatial Index and Its Applications in GIS

空间索引技术可提供高效的空间数据组织与管理方式,以支撑海量空间数据的挖掘与分析。针对当前空间索引存在的知识体系不明晰、选择难等问题,该文通过文献调查法和CiteSpace工具,依据空间划分及映射方法将空间索引划分为基于树结构、格网、空间填充曲线和地址编码的空间索引四大类,并综述其原理、空间结构、适用范围及在GIS领域的应用,最后对空间索引在数据组织、高效计算、可视化、可靠性等方面的研究进行展望。结论如下:基于树结构的空间索引最具普适性且可以处理多维度及多层次的数据,查询性能依赖于树结构的平衡性及数据的分布;基于格网的空间索引可以均匀划分空间以便于高效范围查询,却不适用于非结构化或动态数据集;基于空间填充曲线的空间索引可以在实现维度压缩的同时保持局部邻近性,但插入或删除数据可能导致整个曲线的重构难以频繁更新;基于地址编码的空间索引将语义地址信息转化为编码信息,便于高效检索,然而语义地址匹配仍存在较大误差和不确定性。研究结果可为空间数据组织和结构设计提供参考。

Spatial index is a data structure arranged in a specific order according to the shapes,positions and spatial relationships of spatial objects.This structure can provide an efficient way of data organization and management method to support the mining and analysis of massive spatial data.In view of the problems existing in spatial index,such as unclear knowledge system and difficulty in selecting spatial indexes,with the help of literature survey method and CiteSpace tool,spatial index is divided into four categories by spatial subdivision and mapping methods:tree structure,grid,space-filling curves and address coding.This paper reviews their principles,spatial structures and scope of applications in detail,and provides an outlook on the research of spatial indexes in data organization,efficient computing,visualization,reliability and so on.The conclusion is as follows:spatial indexes based on tree structure are the most versatile and can handle multi-dimensional and multi-level data,while their query performance depends on the balance of the tree structure and the distribution of data;spatial indexes based on grid can evenly divide the space to facilitate efficient range query,but are not suitable for unstructured or dynamic data sets;spatial indexes based on space-filling curves can achieve dimensional compression while maintaining local proximity,but inserting or deleting data may make the reconstruction of the entire curve hard to update frequently;spatial indexes based on address coding can convert semantic address information into coded information to facilitate efficient retrieval,however,there are still large deviations and uncertainties in semantic address matching.The research findings can be used as a reference for researchers in spatial data organization and structure design.