多核环境下Hilbert曲线划分简单要素多边形合并算法

为了解决大规模复杂多边形数据合并运算效率问题,减少在空间数据库中检索多边形时的磁盘读取次数,针对多核环境下简单要素类多边形合并的特点,利用Hilbert曲线划分方法对简单要素的多边形进行空间数据划分,利用多核处理器并发执行,充分利用了多核计算环境中CPU的计算能力,保证了合理的任务分配与充分利用。介绍了该算法中用到的简单要素类多边形合并算子,利用对重合边的判断来进行多边形的合并;最后对提出的算法进行了实验分析。实验证明,本算法在进行大数据的多边形集合合并时效率较高,基于本算法开发的功能用于实际问题中可较好地解决大规模复杂多边形数据层合并运算的效率问题。

Algorithm for simple data model polygon merging based onHilbert-curve on multi-core CPU

For solving efficiency problem of large scale complex polygon data merge and decreasing the number of disk-reading while retrieving polygons in spatial database, this paper considered the charactistics of simple element polygon merging in mult-core environment, divided spatial data with Hilbert curve method, made well useful of computing capability of multi-core environment, priserved reasonable task assigments and used of CPU. And it introduced simply elements polygons merge operator, made use of judgment to coincide to the edge. At last, it carried the experiment analysis. Experiments show that this algorithm has high efficiency while merging large data polygon collection. Functions based on this algorithm can well solve the efficiency of operations, bring by large-scale complex polygon data.