Dynamic Generalization Methods Of POI For The Representation Of Mobile Map

With the development of wireless communication technology and the widespread application of smart phone operating systems, the number of mobile devices (which include cell phones, tablets, and etc.) has exceeded that of PCs. People are more and more fond of getting and sharing information via mobile phones, which brings about more needs for geographic information services and location based services. Driven by social requirements and supported by the software and hardware environments, the mobile maps have been evolving and became an indispensable tool for users in spatial cognition and navigation. POI is an important content in mobile maps and its spatial positioning accuracy, attribute information abundance and representation clarity directly influence the quality and usability of mobile maps. Currently main mobile map service providers are using a strategy which stores data of as many as20levels of detail in multiple resolution spatial databases. Because the quantity of POI is quite large and they mainly distribute beside the road, when they are searched by users, the results showed on the maps are quite unsatisfactory and POIs are overlapping severely. Map generalisation is a potential way to solve these conflicts, but the context of the traditional map generalisation is specified, the target scale, generalisation area and map types are certain and the generalisation model is static, whereas the user requirements of POI in mobile environment are variable. Thus a dynamic system should be built to solve these problems. To design such a system, several aspects must be taken into consideration, such as selection of generalisation algorithms, formalisation of generalisation knowledge and modelling of generalisation process. Details are discussed in the thesis and main works are as follows:1. This paper analysed the representation characteristics of the mobile maps and the POIs and pointed out the drawbacks of current representing methods. Based on the user requirements for POIs in the mobile maps, this research figured out the tasks of POI generalisation and the goals should be achieved. Also the requirements for algorithms in POI generalization were anaylised.2. The existing generalisation algorithms for point feature are reviewed and classified into five categories according to the generalisation operators they belong to. The availability of operators for POI generalisation in mobile environment is analysed from the perspective of keeping generalisation result meaningful and keeping POI and background data match correctly. And the availability of algorithms is anaylised from the perspective of algorithm complexity and characteristics.3. This research explored the existing map generalization models and process controlling methods and explained the reason why they can’t be simply immigrated into mobile environment. Based on the representing characteristics of mobile maps and POIs, a dynamic generalisation model for POI generalisation in mobile environment was proposed. The generalisation knowledge and process controlling method used in this model were discussed.4. Based on the model proposed and the algorithm analysis, a prototype of POI generalisation was designed and implemented with the help of ESRI developing tools. A selection algorithm based on attributes and the circle growth algorithm were implemented to execute the dynamic generalisation process. The result of this experiment proved the feasibility of the dynamic generalisation model proposed.