基于Voronoi图的测度点状目标空间分布特征的方法

Voronoi diagram to study the spatial distribution pattern of point sets

点集的V orono i图是对点集的一种空间分割方式,不同分布的点集可以形成V orono i多边形面积的不同变化,可以通过计算点状目标的V orono i多边形面积的变异系数(CV值)方法,来分析点状目标的空间分布特征.基于国家资源环境数据库,进行了安徽省农村居民点的空间分布特征分析.结果表明安徽省农村居民点总体上属于集群分布,而各县市的分布规律是,中北部平原、低丘陵地区农村居民点呈随机分布,南部、西部山区呈集群分布.最近邻点指数和CV值的相关系数表明,两者之间存在显著的负相关关系.这表明,计算点状目标的V orono i多边形面积的变异系数方法,也是测度点状目标空间分布的一种简便有效的方法.

To define the spatial distribution pattern of point objects, a new approach is introduced, which employs the variation coefficient CV of the area of Voronoi polygons. Based on the National Resources and Environmental Database, this paper studies the spatial distribution pattern of rural settlements in Anhui province, the central China. The result shows that, generally, the spatial distribution pattern of all rural settlements in Anhui province, belongs to the clustered distribution. However, the spatial distribution pattern of those rural settlements, at the central and northern parts af Anhui province, is in the random distribution. The possible reason may be that those rural settlements are at the regions of the plain or low hills, the factors to in.fluence their spatial distribution are not too strong to find their ideal sites freely. On the other hand, at the western or southern parts of Anhui province, because of the terrain effects at mountain area, the locations of rural settlements only be at the sites, which have better natural conditions, such as enough water and arable land, and so forth. And so, their spatial distribution pattern is the clustered distribution. Compared with the method of using the nearest neighbor index, the strong correlation between the nearest neighbor index and CV value shows the method based on the CV of Voronoi polygons＇ area is a feasible and efficient one.