不同耕地空间聚类算法的对比分析研究  

Comparative Study on Spatial Clustering Algorithm of Different Cultivated Land

作　　者：詹秀眉 李亚星[1,2] 欧阳俐 朱文婷 ZHAN Xiumei;LI Yaxing;OUYANG Li;ZHU Wenting(Department of Geography,Minjiang University,Fuzhou 350108,China;Fuzhou Geographic Information Industry Innovation Center,Fuzhou 350108,China)

机构地区：[1]闽江学院地理科学系,福建福州350108 [2]福州市地理信息行业创新中心,福建福州350108

出　　处：《测绘与空间地理信息》2019年第1期112-115,共4页Geomatics & Spatial Information Technology

基　　金：国家自然科学基金项目(41601601);福州市科技项目(2013-S-109);国家级大学生创新创业训练计划项目(201710395006)资助

摘　　要：开展耕地科学分区对于耕地管理利用、种养结合、耕地集约化等方面具有重要的现实意义。本文基于CURE、K-medoid、Single link和BIRCH 4种空间聚类算法,以上街镇耕地数据为例,利用SuperMap iObjects. NET8C组件开展耕地空间聚类对比分析研究,并采用轮廓系数,从凝聚度和分离度的角度比较各种耕地聚类算法的合理性,进而得到耕地的最佳空间聚类结果,以期为区域耕地资源科学配置和优化提供科学依据。It is of great practical significance to cultivate arable land for the management of cultivated land,the combination of planting and breeding,and the intensification of cultivated land.In this paper,using the CURE,K-medoids,Single link,BIRCH spatial clustering algorithm and the above-mentioned cultivated land data as an example,the SuperMap iObjects.NET 8C component was used to study the spatial clustering of cultivated land,and the contour coefficient was used.And the degree of separation is compared with the rationality of various farmland clustering algorithms,and then the best spatial clustering results of cultivated land are obtained,so as to provide scientific basis for scientific allocation and optimization of cultivated land resources.

关 键 词：空间聚类算法 耕地 组件GIS 轮廓系数 

分 类 号：P208[天文地球—地图制图学与地理信息工程]