节点文献
面向“数字海洋”无级比例尺的关键技术研究
Key Technology Research of Infinite Scale on Digital Ocean
【作者】 李一凡;
【导师】 赵德鹏;
【作者基本信息】 大连海事大学 , 交通信息工程及控制, 2002, 博士
【摘要】 “数字地球”是对地球多分辨率的三维表达,它允许我们在不同的视点高度观察地球时可以获得不同详细程度的地理信息,为解决无限的空间地理信息与有限的数据显示窗口之间的矛盾,需要“数字地球”无级比例尺技术。 无级比例尺功能的实现,主要是依靠空间数据库功能来完成,所以本文提出一个面向无级比例尺空间数据库的建立方法。从数据库模型的确立,到数据结构的定义,从数据库提供的功能研究,到数据共享的方式,最后是空间信息的发布。空间数据模型确定了空间数据库能容纳什么样的数据,并决定能提供什么样的服务。空间数据库功能是数据库的关键部分,数据库内部的数据存储与管理,外部用户如何使用这些数据,都是靠功能来实现的。空间数据共享是在数字地球前提下提出的,拥有较好的数据共享方法,数据库才能充分利用现有的数据进行整合。空间数据发布,是将数据库中的数据最后提供给用户的接口,数据只有对外发布才真正具有价值。 空间数据模型和空间数据结构,基本上确定了空间数据库能实现什么样的功能。由于无级比例尺空间数据库应该包含矢量数据和光栅数据,使这两种数据的优缺点互补,所以我们设计的空间数据库数据模型是矢量与光栅一体化的。确定了数据模型,然后确定数据结构。超图结构不仅提供了面向对象的层次结构,而且还提供了层与层之间的联系,非常符合地理空间数据的继承性和关联性。S-57国际标准是国际海道测量组织和国际海事组织有关电子海图数据传输的标准,该标准从理论模型到数据结构定义以及物标的具体定义,本文采用S-57的数据模型,我们采用超图数据结构,建立了一个面向无级比例尺空间数据模型。 制图综合技术为无级比例尺的前提,其后的运算必须要求制图综合完成图形形变,综合时间的判断和综合后形变等。本文采用粗糙集的理论,将不同比例尺下光栅格作为等价划分,给出了平面拓扑上、下近似及边界算法,对拓扑的形变进行研究,提出了拓扑形变算法。同时利用集合相交判断综合时机,并提出临域概念,使得综合时间判断更加灵活。该判断算法较传统算法复杂度降低,计算效率大大提高。利用粗糙集下近似算法,提出了拓扑图形综合后的形变算法。 信息综合作为制图综合的延伸,因为随着图形的变换,图形上所隐含的信息也发生了改变。本文主要讨论了属性信息的综合问题。在属性信息综合问题中,我们提出建立一个空间层次库,由此来判断空间物标的子物标,并将子物标属性信息综合到父物标上。由于每一个物标是一个数据对象,在建立原始空间数据库时,将各个空间物标综合函数确定好,当比例尺到达一定阈值触发相应的综合函
【Abstract】 "Digital Earth" is the 3D expression of Earth multi-resolution, it allow us get geographic information in different details from different viewpoints. In order to resolve the contradiction between infinite spatial geographic information and limited data display window, we need infinite scale technology.The realization of infinite scale mainly depend on spatial database, so the building approach of infinite scale based spatial database is referred in this paper. This approach include the establishment of database model, the definition of data structure, the study of database function, the data sharing method and the release of spatial information. Spatial data model determines which kind of data can be included in spatial database, and which kinds of service are provided. Data storage and management, and the external user how to access data are all depend on the functions provided by database, so spatial database functions is the key part of the spatial database. Spatial data sharing is one of the most important parts of digital earth, and database can take the full advantage of existing data only after having well data sharing mechanism. Spatial data release is the interface which can provide useful data to user, without release, data will have no use.Spatial data model and spatial data structure basically determine the functions of spatial database. To patch the deficiency of vector data and raster data, the data model in our spatial database integrates these two kinds of data. Hypergraph does not only provide the object oriented hierarchy, but also provide the relations among layers. This feature makes it meet the inheritance and relevance of geographic spatial data. S-57 international standard is the electronic chart data transfer standard of International Hydrographic Organization (IHO) and International Maritime Organization (IMO). This standard provides the theoretical model, definition of data structure and concrete definition of objects. This paper adopts S-57 data model and hypergraph data structure to establish an infinite scale oriented spatial data model.Map generalization is the precondition of infinite scale. It includes graph deformation, the judgment of generalization time and the deformation after generalization. To resolve these problems, this paper adopts rough set. We consider raster grid in different scales as equivalence partitioning, and using algorithm of up/down approximation and boundary to study the topological deformation, and bring forward the topological deformation algorithm. Then using aggregation interest to judgegeneralization moment, and bring forward the concept of neighborhood, to make the judgment of generalization moment more flexible. It reduces complexity of traditional algorithm and improves the compute efficiency. According the down approximation algorithm, I bring forward the deformation algorithm after topological graph generalization.Information generalization was treated as the extension of map generalization. With the change of graph, the attributes in these graphs are changed at the same time. This paper mainly discusses the information generalization. In information generalization, we establish a spatial layer database, to judge the sub-objects of spatial objects, and generalize the attributes of sub-objects into parent objects. Each object is a data object, and all kinds of spatial objects generalization functions were established in original spatial database. When scale exceeds the relevant threshold value, it triggers the relevant generalization function. After generalization, it searches in spatial layer database, generalize parent object’s attribute.Spatial data sharing and interchange is another important research aspect of infinite scale database. Firstly, we establish a Meta data oriented management system, and bring forward infinite scale oriented spatial Meta data catalog. The Meta data can describe spatial data and finish the spatial data sharing. According to study the four level of spatial information sharing, we bring forward sharing approach for each level. They are TCP/IP in network level, XML transfer standard in data level, OGC semantic criterion in application level and spatial information frame in enterprise level. All these are based on Internet technology.Spatial information release adopts multi-platform architecture, and using JAVA program browser plug in unit. It defines the concept model and external model of database, defines message transferred between browser and server, and finish the functions in browser, such as zoom, pan, query. In order using the compute ability of workstation, the system equalizes all functions in server and browser.
【Key words】 Infinite scale; Spatial database; Rough set; Map generalization; Information generalization; Spatial data sharing; Spatial information release;