【作者】 田文文；

【导师】 朱欣焰；

【作者基本信息】 武汉大学 ， 地图制图学与地理信息工程， 2013， 博士

【摘要】 随着互联网技术的不断发展,公众越来越多地参与到互联网信息的共享中,用户创建内容(User-generated Content, UGC)成为互联网信息中的重要内容。同时,带有GPS功能的移动终端的广泛应用和地理信息知识的普及也为公众用户自主创建的地理信息进行共享提供了良好的环境。自发地理信息(Volunteered Geographic Information, VGI)正是在这样的环境中诞生的,这种以大众用户网络协作和数据贡献为核心的地理信息,不但为普通人提供了操作和使用地理数据的权利,而且可以用来维护动态更新的开放地理数据库。近年来,基于用户自主创建、公众贡献的自发地理信息网站和应用如雨后春笋般涌现,如维基地图、Google Map Maker以及OpenStreetMap等,其平台以良好的用户使用体验和丰富的数据迅速积累了大量用户,这些数据因为具备更新速度快,语义信息丰富,细节翔实的特点也被各研究机构和行业所广泛应用。自发地理信息的优势在于其更新迅速,可以作为指导专业地理信息数据更新的数据来源之一。自发地理信息来源于公众贡献,专业设备和专业知识的缺乏使得VGI数据采集和制作中可能会带来数据质量问题。传统专业化测绘基于野外实测和遥感影像数据提取的内外业相结合的更新作业方式,其更新数据标准要求严格,但时效性和更新周期受限,缺乏较丰富的语义细节信息。本文的研究重点在于如何将自发地理信息作为专业测绘部门发现变化和数据更新的有效数据来源。基于以上认识,本研究从空间数据变化快速发现、提取与及时更新的目的出发,针对自发地理信息中变化数据的发现、提取、有效性验证和多库更新关联等环节存在的问题,探索基于VGI数据的空间数据变化发现与更新方法和关键技术。本文的研究内容主要包括以下几个方面：1、对目前国内外主流VGI项目的平台发展历程、运作机制和现有的专业测绘部门更新模式进行了深入分析,归纳总结自发地理信息数据类型与特点,研究了自发地理信息作为专业测绘部门的更新来源和依据所存在的数据有效性等问题,提出了基于自发地理信息的空间数据变化发现与更新框架。2、分析了现有的矢量数据同名要素匹配的方法,针对自发地理信息矢量数据中城市路网数据的特点,提出了矢量要素多层次蔓延匹配算法,用于及时发现自发地理信息中的变化数据。通过语义特征得到初始的蔓延单元,结合道路弧段和结点的语义、几何和空间拓扑信息,进而蔓延到整个道路网,直到所有道路弧段被标记,充分顾及了数据之间的互相匹配的关系。3、分析了城市路网矢量数据和高分辨影像中存在的特征知识,提出以获得的VGI更新数据为先验知识,快速捕获变化区域,以从VGI数据中获得的变化地物和现势的高分辨率影像中特征知识为指标,基于模糊推理的置信度计算验证数据有效性。4、根据专业测绘生产部门的实际需求,基于传统生产部门更新机制和数据库组织结构,对在分布式协同环境下多数据库联动更新问题深入分析,提出了基于更新规则的多库联动更新方法,制定多数据库之间的更新规则,实现VGI变化数据从入库、对象化到发布的增量数据更新。5、根据本文提出的基于自发地理信息的空间数据变化发现与更新方法,构建原型系统,对其中用于VGI数据的矢量要素多层次蔓延匹配算法、基于高分辨率影像的VGI数据有效性验证和基于更新规则的多库联动更新方法进行了实现,并在城市路网场景进行了实证研究。基于本文提出的方法,探索了自发地理信息作为专业测绘部门数据更新来源的可能。自发地理信息作为一种有别于传统测绘与遥感的专业地理采集数据,汇聚公众用户的知识,现势性高,细节丰富,在交通、灾害应急等行业具有很大的应用价值。在之后的研究中,我们将根据自发地理信息的特点,关注于数据质量,分析基于公众的数据有效性验证,实现变化数据从识别到验证的一体化,从而构建更为有效的空间数据更新体系。更多还原

【Abstract】 As the development of the Internet technology, citizens are increasingly involved in the Internet information sharing process. Now, the user-generated content is becoming an important source of the Internet information. The widely applications of mobile devices and the popularization of GIS knowledge in citizens are making good foundation to the sharable GIS in UGC. Then VGI (Volunteered Geography Information) emerged under this environment, the spirit of VGI is that users interoperate on the Internet and offers data voluntarily, which means everyone is authorized to operate the GIS data and they can update VGI in the open geospatial data bases dynamically as well.In recent years, a lot of volunteered GIS web sites and applications based on user-generated have been developed, like Wikimapia, Google Map maker, and OpenStreetMap, etc. These websites and applications have been extracted many users because of the good user-experience and abundant data. Addtionally, the features of volunteered data represent on-time data updating, rich sentimental information, and detail descriptions, therefore the data can be applied in the research and industries.The advantage of VGI is to update the data fast and can be used as data source of guiding professional GIS data updating VGI data comes from the public contributes, the lack of professional equipments and knowledge could probably bring about data quality issues during VGI data gathering and production. Traditionally, professional survey combined outdoor surveying and remote sensing to update the data, the standards of updating are more strict, but the timeliness and updating cycle are limited and are short of abundantly sematic detailed information. This paper mainly emphases how to make the VGI as an effective data source of professional survey department to detect the changed for updating. From all of the above, this study will start out at the quick detecting the change, extracting and updating the spatial data, studying the issues of VGI change information in the discovering, extraction, validation, and multi-database correlation in updating process, to explore the method and key techniques of spatial data change detection and updating based on the VGI data. Our research can be summarized as below:Deeply analyzed the development history and operation mechanism of mainstream VGI project platform and the data updating model of current professional surveying department both in domestic and aboard, this paper summarized the data types and characteristics of VGI data, discussed the issues like the data availability of applying VGI data as data updating source in professional survey department, proposed the spatial data change detection and updating framework based on VGI.Analyzed the current homonym element matching method of vector data, this paper proposed the multi-level spreading and increment recognition algorithm based on the characters of city road network in the vector data of VGI. We extracted the highest level semantic features from the matched data to obtain the initial spreading unit, this unit would be matched with the corresponding feature in the referenced data, the unit will spread to the whole road network by matching all the the semantic, geometry, and spatial topology information in the road arcs and nodes until all the road arcs are marked.Analyzed the various knowledge that existed in the vector data of city road network and the high-resolution images, this paper applied the changed data from VGI as priori knowledge, quickly captured the changed area of image, validated the reliability of the changed data by comparing the meta data and feature information of the VGI data and the latest high resolution remote sensing images.According to the actual demands of GIS data production department, this paper proposed data updating rules in multi-databases, which can be achieved a linked data updating in multi-databases under a distributed environment. We designed and implemented the prototype, achieved more efficient incremental data updating with an overall process of storing, objecting and publishing of changed data.Based on the method above, we can solve the existing problems in the VGI as data updating source for professional survey department. VGI is based on the public user’s knowledge about geographic information, which is different from the professional data collected by traditional survey and mapping method, also have high currency and rich details. This kind of data is very valuable in the industries like transportation, emergency response which have high requirements about the timeliness and details of data. In the future research, we will pay more attention to the data quality, analyze the validation of VGI data in the public audit method, and achieve the overall process from detection to validation of data quality, then build a more efficient geospatial data updating system.更多还原