测绘学报 ›› 2016, Vol. 45 ›› Issue (1): 119-126.doi: 10.11947/j.AGCS.2016.20140463

• 地图学与地理信息 • 上一篇    

多边形间空间关系查询的异构多核架构并行算法

谢传节1, 龙舟1,2, 马益杭1,2, 由志杰1,2   

  1. 1. 中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室, 北京 100101;
    2. 中国科学院大学, 北京 100049
  • 收稿日期:2014-09-04 修回日期:2015-05-04 出版日期:2016-01-20 发布日期:2016-01-28
  • 作者简介:谢传节(1971—),男,博士,副研究员,研究方向为并行地理计算。
  • 基金资助:
    国家863计划(2011AA120302;2011AA120306;2012AA12A401);海洋公益性项目(201105033-6)

Parallel Algorithm of Spatial Relationship Query between Polygons Based on Heterogeneous Multi-core Architecture

XIE Chuanjie1, LONG Zhou1,2, MA Yihang1,2, YOU Zhijie1,2   

  1. 1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China;
    2. University of Chinese Academy of Sciences, Beijing 100049, China
  • Received:2014-09-04 Revised:2015-05-04 Online:2016-01-20 Published:2016-01-28
  • Supported by:
    The National High-tech Research and Development Program of China (863 Program) (Nos.2011AA120302;2011AA120306;2012AA12A401);The Marine Public Welfare Project (No.201105033-6)

摘要: 目前在空间关系查询中常用的Plane Sweep算法是一种串行算法,在处理海量空间数据时效率较低,而已有的并行计算方法对于普通的计算机并不适用。本文针对这个问题,提出了一种多边形间空间关系查询的异构多核架构并行算法,该算法先利用STR树索引过滤掉不相交的多边形,然后将过滤后的多边形数据集合分解为点集合和边集合,并对其构建四叉树索引;在保证数据浮点运算精度符合要求的情况下,利用GPU强大的批量运算能力快速处理边与边的相交情况并据此逐步计算得到环间的拓扑关系,再根据环间拓扑关系计算得到多边形间的维度扩展九交模型(DE-9IM)参数值;根据DE-9IM参数值与空间关系查询条件相比对,输出查询结果。最后通过试验验证了算法的准确性与高效性。

关键词: 异构多核, 并行计算, 拓扑关系, 空间关系查询

Abstract: The commonly used Plane Sweep algorithm in the spatial relationship query is a serial algorithm, when dealing with huge amounts of spatial data, the efficiency is very low,and the existing parallel computing method is not applicable to normal computer. Aiming at this problem, this paper proposes a parallel algorithm of spatial relationship query between polygons based on heterogeneous multi-core architecture. The algorithm uses STR-tree index to filter out non-intersection polygons first, then decomposes the filtered polygons data into points set and edges set, and constructs a quadtree index for them; under the premise of data accuracy meets the requirement of floating point calculation, uses GPU's powerful batch computing power quickly processing the intersection between edges and calculates the topology relationship between rings, then uses the topology relationship between rings to calculate the DE-9IM model between polygons; compares the DE-9IM model with spatial relationship query condition and outputs the query results. At last, the efficiency and accuracy of the algorithm is verified by experiment.

Key words: heterogeneous multi-core, parallel computing, topological relations, querying spatial relationship

中图分类号: