摘要
针对空间连接方法中大部分的研究都是基于拓扑和距离的连接方法,而基于方向关系查询的研究则比较少,并且查询性能较低等问题.本文提出了一种新的方向关系查询精过滤方法,在传统的两步查询处理过程中,增加了一个精过滤步骤,改进后的方法可以达到过滤候选集的大小效果,从而进一步减少求精步骤计算的工作量.通过理论的分析和实验,验证了改进后方法需要进行精算的计算量仅为原有方法的40%左右,充分证明了该方法的有效性.实验结果表明,与传统的方向关系查询处理方法相比,本文方法在CPU的计算时间和I/O方面分别提高了40%和20%左右.
The research work on the processing of spatial joins has primarily focused on topological and distance relations. However, there has been little work on the study of queries for directional relations. In addition, the query performance is low. A new fine query filtering method based on an R-tree was presented. A fine filtering step was added in the middle of the traditional 2-stage query model. The improved method could achieve the effect of the filtering candidate set, so to further reduce the workload of the refining step. The validity of the method has been proven by theoretical analysis. The improved method only needs 40% of the fine calculation workload of the original method, and the results fully prove the validity of this method. Finally, concrete test data was used to prove its validity. The experiment results show that the CPU-time and I/O of the proposed method are improved by 40% and 20% respectively.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2010年第11期1490-1495,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(60873037)
关键词
方向关系
方向连接
空间数据库
空间数据挖掘
空间索引
direction relation
direction join
spatial database
spatial data mining
spatial index
