特征分类与邻近图相结合的建筑物群空间分布特征提取方法

发布时间：2018-07-01 13:35

本文选题：地图综合 + 建筑物群　；参考：《测绘学报》2017年05期

【摘要】：建筑物群综合过程中需要对建筑物群空间分布特征进行认知和识别。本文在分析国内外相关研究的基础上,从描述建筑物空间特征的大量指标中,利用主成份分析方法,总结并提出了有代表性的建筑物空间特征指标集:凸包面积、紧密度IPQ指标、边数和最小面积外接矩形方向,并基于这些指标研究了建筑物群的分类。在利用最小生成树邻近图(MST)划分建筑物空间子群时,考虑了建筑物成群与所处地理环境(河流和道路等因素)的关系。另外,基于最邻近图(NNG)、MST、相对邻近图(RNG)和Gabriel图(GG)4种建筑物群邻近图,提出了自动识别具有特定空间排列建筑物子群的方法,并比较分析了识别结果的影响因素和可用性。最后,选择北京某地区建筑物群为试验对象,实现了对建筑物群的分类和空间聚类,并提取了其中直线型空间排列的建筑物子群。
[Abstract]:In the comprehensive process of building groups, it is necessary to recognize and identify the spatial distribution characteristics of the building groups. On the basis of the analysis of relevant research at home and abroad, this paper summarizes and puts forward a representative set of spatial characteristics of buildings: the area of the convex hull, and the tight area of the building. The degree IPQ index, the edge number and the smallest area are connected to the rectangular direction, and the classification of the building groups is studied based on these indexes. In the use of the minimum spanning tree adjacent graph (MST), the relationship between the groups of buildings and the geographical environment (rivers and roads and other factors) is considered. In addition, it is based on the nearest neighbor graph (NNG), MST, relative neighbour. Near graph (RNG) and Gabriel diagram (GG) 4 building groups adjacent graphs, the method of automatic identification of building subgroups with specific spatial arrangement is proposed, and the influencing factors and availability of the recognition results are compared and analyzed. Finally, the building group in a certain area of Beijing is selected as the test object, and the classification and spatial clustering of the building groups are realized, and the extraction of the building group is realized. A subgroup of buildings in which the linear space is arranged.
【作者单位】：武汉大学资源与环境科学学院;武汉大学测绘遥感信息工程国家重点实验室;中国测绘科学研究院;
【基金】：国家自然科学基金(41471384) 公益性科研专项(201512032)~~
【分类号】：P208

【相似文献】