【作者】 王刚；

【导师】 杜清运；

【作者基本信息】 武汉大学 ， 地图制图学与地理信息工程， 2011， 博士

【摘要】 随着互联网软硬环境建设的不断发展,互联网因其固有的分布式特征已逐步成为数据发布、数据共享、分布式计算的平台,借助于互联网这一广阔的信息传输平台,GIS应用领域得到了更大限度的拓展,在网络技术,多媒体技术、空间信息技术等的推动下,WebGIS技术得到了更为深层次方面的发展与应用。在GIS应用中主要存在两类数据分别为矢量数据与栅格数据,矢量数据由于数据结构紧凑、冗余度低,有利于网络和检索分析,图形显示质量好,精度高等优点而得到广泛地应用,而栅格数据则凭借着数据结构简单,便于空间分析和地表模拟,现势性较强等优势在目前国土监测、灾害分析、环境监测等方面发挥了重要的作用。目前由于矢量数据应用的广泛性和实用性,使得矢量数据成为空间信息化建设的基石,同时也因为针对不同的应用需求,各种GIS软件商对数据结构和模型采用不同的设计理念,使得矢量数据的数据结构呈现的多样性和复杂性,但目前以上的各种数据模型还不能很好地解决WebGIS中矢量数据的快速传输和客户端显示等问题。随着web2.0技术的发展与成熟,基于栅格数据下的高性能网络传输已经通过分块模型得到了基本的解决,使得以栅格图片格式为基础的WebGIS技术成熟起来,并投入到国家信息化建设的各个领域,但由于栅格数据自身所固有的缺陷,使得目前基于栅格图片技术建立的WebGIS使用受到了功能的限制。因此,基于矢量数据的WebGIS的研究与探索就成了目前关注的焦点话题之一。文章首先从WebGIS基本特征、实现的技术模式以及体系结构等方面详细叙述了WebGIS实现机制,从应用方面分析了各种WebGIS搭建方法的优势和不足,结合目前成熟的基于地图图片的WebGIS技术,通过实验分析了这种技术下数据组织模型、数据交互式传输以及数据展示等方面的,同时也分析当前采用矢量数据渐进式传输的数据组织思想、特点以及存在的问题。其次,本文从缓存的存储方式、服务器端缓存和客户端缓存等方面讨论了如何将这种技术应用到基于矢量数据的WebGIS中,从而最终减少数据的重复性传输；阐述了四叉树数据结构,提出了基于文件存储的线性四叉树构建方法,并讨论了同一等级和上下级之间的拓扑关系；同时结合矢量数据渐进式传输中用到的模型以及WebGIS的理论,采用以空间换时间的思想,提出了基于层次增量分块矢量模型的服务器端矢量数据组织模型,并给出了相应的矢量块的剪裁及文件命名存储等的方法,同时也给出了客户端的矢量块文件数据融合实现机制。接着,本文从与网络传输密切相关的矢量数据数据量入手,从数据压缩的基本原理和一般方法阐述了数据压缩的实质,从局部压缩思想和整体压缩思想两方面对矢量数据五种有损压缩算法进行了阐述与分析,考虑到以上压缩方法存在数据信息丢失的问题,所以本文从基于统计模型和基于字典模型两方面阐述了Huffman编码、算术编码以及基于第一类字典编码、第二类字典编码的无损压缩算法,同时结合矢量数据特点、网络传输以及数据压缩,指出除了要建立一种合适的矢量数据模型外,还要对矢量数据传输前进行多种组合的压他缩,并给出了矢量数据坐标几何压缩的方法以及基于Gzip编码的网络压缩与传输的思路,从而减少了网络传输的矢量数据量；同时结合服务器端的矢量数据组织,提出了从控制刷新数据量,采取基于特征要素交互式传输和基于Web Service的交互式异步传输等方面的传输策略来提高矢量数据的传输效率,最终达到提高用户体验的目的。文章最后通过实验从网络传输数据量、传输时间以及客户端数据显示与编辑等方面对前面叙述的数据压缩、缓存技术运用、传输策略以及数据融合等进行验证,同时也指出以上方式存在的问题。更多还原

【Abstract】 With the development of the hardware and software, the Internet has gradually become the platform of data releasing, date sharing, and distributed computing. Because of it’s inherently distributed features. By means of the Internet and driven by the network technology, multimedia technology and spatial information technology, WebGIS has broad development and applications.In GIS applications, there are two main types of data:vector data and raster data. Vector data has the features of compact data structure, low redundancy, good graphical display quality, and high precision. So vector data are conductive to networking and search analysis. The raster data has the features of simple data structure, easy to surface modeling and spatial analysis, etc. Raster data has played an important role in land monitoring, hazard analysis, environmental monitoring, and so on.At present, because of the extensive and practical applications of vector data, it has become the cornerstone of spatial information construction. In order to meet the different application requirements, a variety of GIS software have their different data structures and business models to make the vector data structure to demonstrate the diversity and complexity. However, above all the data models can not solve the issues of rapid vector data transmission and display in the clients of WebGIS.With the development of Web2.0, the high-performance network transmission of raster data has been solved by the block model. The WebGIS technology based on the raster image formats has become matured and applied in the fields of national information construction. But because of raster data’s inherent deficiencies, WebGIS based on the raster image technology has the restrictions of its function applications. Therefore, the exploration of WebGIS based on vector data has become one of the current hot topics of concern.This paper firstly describes WebGIS implement mechanism in detail from the aspects of its basic features, technical model and architecture, And WebGIS construction methods from the applications. What’s more, combined with current mature WebGIS technology based on map image, the author analyzes not only the data organization model, interactive data transmission, data presentation, but also data organization thinking, characteristics and problems of the current progressive transmission of vector data through the experiments.Secondly, this paper discusses how to apply the cache technology to the WebGIS based on vector data in order to reduce the repeatability of data transmission. Then, the author analyzes quadtree data structure, proposes a linear quadtree construction method based on file storage and discusses the topological relations of the same level and different levels. Combined with progressive transmission model of vector data, WebGIS theory and idea of space for time, the author proposes vector data organization model based on block-level incremental vector model in the server, the corresponding methods of cutting vector blocks and file storage, and the methods of data fusion in the client.Thirdly, this paper describes the essence of the vector data compression and five lossy compression algorithms from the local compression and the overall compression. Taking into account the loss of data compression, this paper analyzes the Huffman coding, arithmetic coding, dictionary coding, Combined with vector date features, network transmission and data compression, the author analyzes the establishment of a suitable vector data model, compression vector data before data transmission, the geometric compression of vector data coordinates, and idea of network compression and transmission based on Gzip coding to reduce the amount of vector data transmission. Meanwhile I propose a new strategy to enhance transmission efficiency of vector data through a combination of vector data organization at server side, reducing update data volumes, interactive transmission based on feature characteristics, as well as interactive as asynchronous transmission methods. By means of these methods, we finally achieve the goal of promoting user experience.Finally, the author verifies the described data compression, cache application, transmission strategy, and data fusion, by the experiments from the amount of network data transmission, transmission time, data display and editing in the client. Furthermore, the author points out the problems as above.更多还原