面向多粒度时空对象数据模型的网络电子地图生成方法

Generation Methods of the Web Electronic Map for Data Model of Multi-granularity Spatiotemporal Object

传统网络电子地图生成是以要素进行组织的,在表达地理实体动态变化、关联关系以及多粒度特征时存在一定的局限性。而多粒度时空对象数据模型旨在解决现实世界到对象所组成的事物空间之间的映射这一科学问题,为时空实体的可视化提供了新的思路,为展示地图要素间复杂关联、多维动态等特征提供了模型基础和数据保证。本文将多粒度时空对象数据模型引入网络电子地图生成当中,渐进得改变了以往网络电子地图依靠图层数据生成的模式,基于多粒度时空对象的概念、模型框架以及数据存储与管理方式,提出了2种网络电子地图生成方法。一种是利用多粒度时空对象数据重建图层,为现有网络制图工具提供“中间件”,使其能够调度和使用多粒度时空对象数据;另一种是将多粒度时空对象数据和地图符号进行绑定,使网络电子地图不再依赖于图层而能根据多粒度时空对象动态更新,同时便于将对象之间的关系、组成结构等非结构化属性特征进行可视化。同时,本文通过实验验证了2种方法的可行性,为多粒度时空对象在网络电子制图领域的应用进行了有效实践。

Traditional web electronic map is a composition of several layers containing geometry elements,which rely on an object-oriented model. However, this model is too abstract to map a more complex world. So the maps based on this model have limitations in expressing the characteristics of dynamic changes and correlation of geographical entities. Domain experts have proposed the data model of Multi-Granularity SpatioTemporal Object(MGSTO) in recent years, aiming to solve problems of mapping the real world to the“thing space”which is full of“objects”. This inspires new visualizations on the spatio-temporal phenomenon, at the same time, changes the way of map making. Visualizations or maps based on this data model have access to data that are more reliable. The problem is that there is no such map making technique or tool that can deal with it.Therefore, we propose two map marking methods that can make full use of this data model. Firstly, we think it is necessary to bridge existing web cartography platforms and the MGSTO database, so we propose the layerreconstruction method. In this way, all the online mapping platforms have the ability to understand the data model of MGSTO and make maps depending on it. This method looks like a "middleware" of online mapping tools. Secondly, we propose a method that binds MGSTO onto HTML DOMs, for example, Scalable Vector Graphic(SVG) objects. The MGSTO data is visualized directly into graphic elements(or symbols) and its changes and transforms can be easily reflected onto symbols. The advantage of this method is that updating maps is synchronized with database updates. Maps made by this method can be linked to the real world due to the data model of MGSTO, which has the ability to manage multi-scale features. More complex information such as relation and composition can be easily extracted from the data model of MGSTO, which can be further mapped as thematic maps. Multi-granularity features can be shown by this method. Each of these two map making methods has distinct characteristics. The first method provide mature and stable style of geographic data visualization, and it can be published online easily because it use online infrastructure provided by online mapping platforms. The second method provide map makers more freedom in map design. Users can perfect the data model of MGSTO targeting certain map making objectives. However, more algorithms such as generalization need to be developed for the data model of MGSTO. In addition, this research carries out two experiments using the same dataset to verify our methods, which demonstrates the feasibility our proposed methods.