一种不同尺度下点群目标相似度计算模型  

Spatial Similarity Assessment of Point Cluster Method in Multi-scale Map Spaces

作　　者：李金玉[1] 张椿玲[2] LI Jinyu;ZHANG Chunling(School of Electronic and Information Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China;Eletrical Engineering College,Northwest University for Nationalities,Lanzhou 730070,China)

机构地区：[1]兰州交通大学电子与信息工程学院,兰州730070 [2]西北民族大学电气工程学院,兰州730070

出　　处：《遥感信息》2017年第6期127-134,共8页Remote Sensing Information

基　　金：甘肃省科技计划项目(1506RJZA084);兰州市科技计划项目(2015-2-74)

摘　　要：为了更为准确地表达不同尺度下点群目标的相似关系,针对前人研究未能顾及点群目标中的空白区域,利用自适应空间聚类方法,对点群目标进行聚类,得到点群目标内部空白区域。通过对空白区域的研究,提出相应的相似性计算模型;找到点群目标相似性的影响因子,包括拓扑关系、方向关系、分布范围、距离关系等,并提出相应的相似度计算方法,最后整合成整体的点群相似度计算模型。该方法能准确表达不同尺度间点群目标相似性的差异,对制图综合的质量评价以及点群目标的检索具有一定的参考价值。In order to express the similarity relation of the point group target in different scales,this paper uses the adaptive spatial clustering method to cluster the target of the point group and get group targets interior blank area,which cannot be taken into account in the previous study.Through the study of the blank area,the corresponding similarity calculation model is put forward.The influencing factors of the similarity of the target groups are found,including the topological relation,the direction relation,the distribution range and the distance relation.Then this paper put forward the corresponding similarity calculation method and the whole point group similarity calculation model.This method can accurately express the difference of the similarity of the point targets at different scales,which has certain reference value for the quality evaluation of the cartographic synthesis and the retrieval of the point group targets.

关 键 词：空间聚类 不同尺度 点群目标 空间相似度 空间关系 

分 类 号：P208[天文地球—地图制图学与地理信息工程]