栅格数据多源成本距离分析并行算法及实现  

Parallel Algorithm for Raster-based Multiple Sources Cost Distance Analysis

作　　者：汪维录[1] 马劲松[1] 于晓艳[1] 

机构地区：[1]南京大学地理与海洋科学学院,江苏南京210093

出　　处：《微计算机信息》2011年第6期221-223,共3页Control & Automation

摘　　要：针对栅格数据成本距离分析串行算法在大数据量、多源点的情况下计算效率低的问题,在串行算法的基础上,设计并实现了一种并行算法。实验结果表明,该算法有效地提高了计算效率。当栅格数据越大,点状源数目越多的时候,并行算法较串行算法的优越性越明显,并且随着线程个数的增加,并行算法的执行效率也在不断提高。In view of the serial algorithm for raster-based cost distance analysis is inefficient in the case of large amounts of data and multiple sources,a parallel algorithm was designed and implemented on the basis of serial algorithm.Experimental results show that this algorithm improves the calculating efficiency.When the raster size and the number of the source point is bigger,the advantage of the algorithm is greater.It also shows that the algorithm performs better if there are more threads.

关 键 词：成本距离分析 多源 并行算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] P208[自动化与计算机技术—计算机科学与技术]