一致性约束下的土地利用狭长图斑降维综合方法研究

发布时间：2018-07-28 15:25

【摘要】：土地利用现状图是指反映一定区域内的土地利用类型、数量、质量及结构的专题地图。与一般的地理数据类似,土地利用数据包含空间数据和属性数据两部分,具有空间数据全覆盖、无重叠、无缝隙及属性数据语义上多层次划分的特点。因此,土地利用数据制图综合有别于传统的普通地图的制图综合,除了需要考虑传统普通地图制图综合的理论因素外,还需要根据土地利用数据特点,考虑其所受到的拓扑一致性约束及语义一致性约束。目前,有很多的学者对土地利用数据制图综合的操作算子进行了大量的研究,取得了一系列的成就,但现有的研究大多集中于土地利用数据中的图斑合并或融合方面,甚少有专门针对图斑降维操作的研究。基于此,本文在前人研究的基础上,考虑土地利用数据自身的特点,深入研究了土地利用数据中图斑由面降维成线的方法。其中主要的研究包括:(1)在空间点集拓扑理论上探讨了土地利用现状图图斑降维成线后所产生的拓扑关系变化,并基于拓扑关系一致性的约束,分析了图斑在不同尺度下,面-面拓扑关系与线-面及线-线拓扑关系的等价性推理;(2)介绍了对狭长图斑的定量化描述方法,并基于约束Delaunay三角网给出了狭长图斑的自动探测及图斑狭长分支的分割方法;(3)研究基于圆扩展的拓扑共享点探测方法,对降维后所产生的不同拓扑关系变化的狭长图斑情况给出了不同的一致性改正方法。同时采取了ArcGIS Engine二次开发的方法在Visual Studio 2010中构建了相应的实验算法模块,实现了文中所提出的方法。最后,本文进行了包括小图斑的合并、狭长图斑的自动探测、图斑狭长部分的分割、图斑自动降维及一致性改正五个实验。实验采用南京市高淳区淳溪镇1:5000 土地利用数据,目标比例尺为1:50000比例尺,实验结果表明综合后的零散小图斑减少,包括公路用地、农村道路用地、田坎、沟渠四种狭长的用地类型被降维成线,降维后的数据依然维持“空间全覆盖、无重叠、无缝隙”的特点,且区域总面积基本维持平衡,反映了所提出的一致性改正算法的有效性。整体实验结果表明了本文所提出的方法的有效性和可行性。
[Abstract]:The current land use map refers to the thematic map reflecting the types, quantity, quality and structure of land use in a certain area. Similar to general geographic data, land use data includes two parts: spatial data and attribute data, which have the characteristics of full coverage of spatial data, no overlap, no gap and semantic multi-level partition of attribute data. Therefore, the cartographic generalization of land use data is different from the cartographic generalization of the traditional common map. In addition to the theoretical factors that need to be taken into account in the cartographic generalization of the traditional general map, it also needs to be based on the characteristics of the land use data. The topological consistency constraint and semantic consistency constraint are considered. At present, many scholars have done a lot of research on the operator of land use data mapping generalization, and have made a series of achievements, but most of the existing research focuses on the integration or fusion of land use data. There is very little research on the dimension reduction operation of pattern spot. On the basis of the previous researches and considering the characteristics of the land use data, this paper deeply studies the method of reducing the dimension of the map spot from the surface to the line in the land use data. The main research contents are as follows: (1) based on the topological theory of spatial point set, the changes of topological relationship caused by the reduction of land use map are discussed, and based on the constraints of the consistency of topological relations, the map spots are analyzed at different scales. The equivalence reasoning between the topologic relation of plane and the topological relation between line and line. (2) the quantitative description method of long and narrow picture spot is introduced. Based on constrained Delaunay triangulation, the automatic detection of long and narrow spots and the segmentation method of long and narrow branches are given. (3) the method of topology sharing point detection based on circle expansion is studied. In this paper, different consistency correction methods are given for the long and narrow picture spots with different topological relationship changes after dimension reduction. At the same time, the method of ArcGIS Engine secondary development is adopted to construct the corresponding experimental algorithm module in Visual Studio 2010, and the method proposed in this paper is realized. Finally, five experiments are carried out in this paper, including the combination of small spots, the automatic detection of long and narrow spots, the segmentation of long and narrow parts, the automatic dimensionality reduction and consistency correction. The experiment uses the land use data of 1: 5000 in Chun Xi Town, Gaochun District, Nanjing City. The target scale is 1: 50000 scale. The experimental results show that the scattered small map spots are reduced, including road land, rural road land, farmland, etc. Four long and narrow land types of ditches are reduced into lines, and the data after dimensionality reduction still maintain the characteristics of "full space coverage, no overlap, no gap", and the total area of the region is basically balanced. It reflects the effectiveness of the proposed consistency correction algorithm. The experimental results show that the proposed method is effective and feasible.
【学位授予单位】：南京师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F301.24;P208

【参考文献】