基于Kohonen网络的点群综合研究

基于Kohonen网络模型,对标准的SOM(self-organizing feature map)算法进行了改进,在保持点群原有空间分布特征的情况下研究点群的选取和典型化。实践表明,该方法适合任意空间分布类型的点群综合。     

Delaunay三角网在图斑与点群目标综合中的应用

土地资源调查利用的土地类型数据多为不同比例尺电子矢量地图,以此数据为研究基础,详细介绍了Delaunay三角网及其Voronoi多边形的特性,以及在土地数据类型中的图斑综合与点群目标综合中的算法应用,具体描述了两种算法的实现过程和试验结果,并对几种点群目标算法进行了简要分析和比对,详细介绍了实现过程;重点阐述了图斑综合算法的思想和算法的具体实现.最后以"土地利用数据库智能化自动综合系统"为例验证了算法方案的合理性和可行性.     

Delaunay三角网在图斑与点群目标综合中的应用

土地资源调查利用的土地类型数据多为不同比例尺电子矢量地图,以此数据为研究基础,详细介绍了Delaunay三角网及其Voronoi多边形的特性,以及在土地数据类型中的图斑综合与点群目标综合中的算法应用,具体描述了两种算法的实现过程和试验结果,并对几种点群目标算法进行了简要分析和比对,详细介绍了实现过程;重点阐述了图斑综合算法的思想和算法的具体实现。最后以"土地利用数据库智能化自动综合系统"为例验证了算法方案的合理性和可行性。     

基于改进Kohonen神经网络的遥感影像分类

提出了一种基于改进后的Kohonen自组织特征映射神经网络的聚类方法,应用于多光谱遥感影像分类处理,试验证明其分类精度有较大提高。     

基于CIRCLE特征变换的点群选取算法

本文在分析了当前地图自动综合算法的基础上,提出了一种基于Circle特征变换的点群选取新算法。文中介绍了该算法中空域中心点的确定、目标特征空间的计算、坐标空间到特征空间的转换、特征空间的聚类以及特征空间的化简等关键步骤。最后就算法的结果进行了分析与评价。     

凸壳原理在点群目标综合中的应用

对聚焦分布的点群，借助凸壳算法形成多层嵌套，以反映它的逐层分布特征，为点群目标的２经选取提供整体分布控制，借助ＶＯＲＯＮＩ图各个特体的区域性评价提供补充性的量化依据。综合过程分为两个子过程：凸壳层的合并和多边形折线顶点的综合。     

基于CIRCLE特征变换的点群选取改进算法

本文在介绍了已有的“基于Circle特征变换的点群选取算法”(测绘科学2005年第3期)基础上,提出了一种改进算法。详细分析了已有算法在线化简过程中对聚类的过度依赖性和线化简过程中的逻辑不一致性等两个主要不足,对特征空间中的线化简等关键步骤进行了改进。最后就改进算法的结果与原先算法的结果进行了对比分析与评价。     

两种地图点群综合算法的比较

通过搭建实验程序,分别对基于非加权Voronoi图和加权Voronoi图的点群自动综合算法进行验证,详细对比和分析两种算法计算结果对点群的统计信息、拓扑信息、度量信息和专题信息的传递情况,计算部分结果的相似度,从而得出加权Voronoi图点群综合算法更优的结论。     

利用层次Voronoi图进行点群综合

通过距离权重描述点的重要程度,采用改进的k-means算法得到点群的聚类中心,进而以聚类中心为基础,构建了层次加权Voronoi图与Voronoi层次树结构.以点群的分布范围、排列方式与密度为度量,给出了基于Voronoi层次树结构的点群综合方法,确保了点群综合前后在空间形态分布上的一致性.结合地理统计学计算,对综合方法作了进一步的量化评估与优化.经验证,本文方法是可行、有效的.     

面向特征的网络空间点群要素自动综合方法

网络空间信息可视化对揭示网络空域规律、促进网络空间认知具有重要意义.将网络空间节点与拓扑关系直接可视化的视图中存在大量的点重合和线交叉,目前已有的网络节点布局算法、集束边技术、骨干网提取和网络路由拓扑多尺度表达等方法能够优化视图效果,但在网络的微观结构上,对保持网络空间点群要素的特征信息关注不够.通过分析并量化网络空间...     

Point set generalization based on the Kohonen Net

Point set generalization is one of the essential problems in map generalization. On the demands analysis of point set generalization, this paper proposes a method to generalize point sets based on the Kohonen Net model; the standard SOM algorithm has been improved so as to preserve the spatial distribution properties of the original point set. Examples illustrate that this method suits the generalization of point sets.     

Schematic road network map progressive generalization based on multiple constraints

Multiple constraints for schematic road network map cartographic design are analyzed and summarized. Based on this, a set of quantitative criteria are set up and a new road network generalization method including progressive selection and displacement is proposed. Furthermore, topological checking methods for road networks are researched. Based on these constraints, the points in a road network are classified, and a satisfactory and effective schematic map is designed in a concrete experiment while maintaining topological consistency of the road network between the original and the schematic map.     

多约束条件下示意性道路网地图渐进式综合

Multiple constraints for schematic road network map cartographic design are analyzed and summarized. Based on this, a set of quantitative criteria are set up and a new road network generalization method including progressive selection and displacement is proposed. Furthermore, topological checking methods for road networks are researched. Based on these constraints, the points in a road network are classified, and a satisfactory and effective schematic map is designed in a con- crete experiment while maintaining topological consistency of the road network between the original and the schematic map     

Subdivision of polygon parcel in land-use data generalization

In land-use data generalization, the removal of insignificant parcel with small size is the most frequently used operator. Traditionally for the generalization method, the small parcel is assigned completely to one of its neighbors. This study tries to improve the generalization by separating the insignificant parcel into parts around the weighted skeleton and assigning these parts to different neighbors. The distribution of the weighted skeleton depends on the compatibility between the removed object and its neighbor, which considers not only topological relationship but also distance relationship and semantic similarity. This process is based on the Delaunay triangulation model. This paper gives the detailed geometric algorithms for this operation.     

Subdivision of polygon parcel in land-use data generalization

In land-use data general-ization, the removal of insignificant parcel with small size is the most fre-quently used operator. Traditionally for the generalization method, the small parcel is assigned completely to one of its neighbors. This study tries to improve the generalization by sepa-rating the insignificant parcel into parts around the weighted skeleton and assigning these parts to different neighbors. The distribution of the weighted skeleton depends on the com-patibility between the removed object and its neighbor, which considers not only topological relationship but also distance relationship and semantic sim-ilarity. This process is based on the Delaunay triangulation model. This paper gives the detailed geometric al-gorithms for this operation.     

大比例尺数字地形图缩编方法的探讨

本文针对大比例尺地形图全自动化制图综合所存在的难点,分析了影响制图综合实现的主要因素.通过生产实践提出基于规则库的一致性制图综合和交互式制图综合相结合的方法,提高计算机的人工智能水平,从而形成完善而高效的制图综合工作流程.实践证明,基于规则库的一致性制图综合和交互式制图综合方法作业效率比较高,可以顺利地完成地图缩编工作...     

顾及局部特征的线状要素制图综合

针对传统的线状要素化简方法难以准确描述线要素局部结构特征的问题,该文提出了一种顾及局部结构特征的线状要素制图综合方法。引入m阶邻居坐标点概念,分析坐标点对线要素局部结构特征的贡献程度;建立了局部结构特征度量方法,实现了简单线要素的自动化简;基于该化简过程,从拓扑不变性角度顾及线状要素的整体形态及不同线状要素之间的拓扑关系,实现了复合线划网络的自动综合。基于对简单线划目标和复杂线划网络的实验结果表明:通过调整m的取值,可以完成不同尺度下线要素的自动制图综合,且化简结果较好地保持了原始线划要素的整体形态特征。     

道路综合中基于骨架分割的道路密度分析

This paper proposes an algorithm for road density analysis based on skeleton partitioning. Road density provides metric and statistical information about overall road distribution at the macro level. Existing measurements of road density based on grid method, fractal geometry and mesh density are reviewed, and a new method for computing road density based on skeleton partitioning is proposed. Experiments illustrate that road density based on skeleton partitioning may reveal the overall road distribution. The proposed measurement is further tested against road maps at 1:10k scale and their generalized version at 1:50k scale. By comparing the deletion percentage within different density interval, a road density threshold can be found, which indicate the need for further operations during generalization. Proposed road density may be used to examine the quality of road generalization, to explore the variation of road network through temporal and spatial changes, and it also has future usage in urban planning, transportation and estates evaluation practice. Supported by the National 863 Program of China(No.2007AA12Z225), the Natural Science Foundation of China(No.40771168).     

Frameworks for generalization constraints and operations based on object-oriented data structure in database generalization

1　ＩｎｔｒｏｄｕｃｔｉｏｎＤａｔａｂａｓｅｇｅｎｅｒａｌｉｚａｔｉｏｎｃａｎｂｅｃｏｎｓｉｄｅｒｅｄａｓｔｈｅｔｒａｎｓｆｏｒｍａｔｉｏｎｏｆｔｈｅｃｏｎｔｅｎｔｏｆａｓｐａｔｉａｌｄａｔａｂａｓｅｆｒｏｍｈｉｇｈｒｅｓｏｌｕｔｉｏｎ (ｗｉｔｈｍｏｒｅｄｅｔａｉｌ)ｔｏａｌｏｗｅｒｒｅｓｏｌｕｔｉｏｎ (ｗｉｔｈｌｅｓｓｄｅｔａｉｌ)ｔｅｒｒａｉｎｒｅｐｒｅｓｅｎｔａｔｉｏｎ(Ｍｏｌｅｎａａｒ ,1 996 ) .Ｉｎｏｔｈｅｒｗｏｒｄｓ ,ｔｈｉｓｔｒａｎｓｆｏｒｍａ ｔｉｏｎｉｓｄｅｅｍｅｄａｓｃｈａｎｇｉｎ…