Research Of 3D Model Data Management Technology For Virtual City

The construction and application of 3D city has become an inevitable trend of the development of information technology and it is an import part of "Digital City" construction. The visualization of virtual city will need lots of 3d model data built for the city landscape. The way how to organize and manage such massive data is the most basic and also import issue, it will affect the performance of data visualization directly. This article will focus on the management mode of 3d model data and try to build an efficient and solid model data service.In model data object organization, a traditional way is taking the full model file as a unit for data object, considering the feature of 3d model data (large amount of data with high data-density) and the scheduling requirements of virtual city visualization applications, this paper construct a smaller model data unit whose data structure is also fit for visualization. This way of data organization can increase the flexibility of scheduling and also avoid the problem of data formatting for visualization.In data management, usually, there are two different ways which are data management based on spatial database and data management based on file system. In this paper, we design and implement those two ways. For the former way, we implement the design of database mode considering the description of CityGML about city object and the demand of data scheduling which makes the structure of model object more reasonable and it works well for the management of data collection in a small area. For the latter way, we expand it to distributed file system to satisfy the growth of data collection in future. Based on the locality feature of data scheduling we organize data based on grid build a secondary spatial index system based on grid, in this way, we can reduce the data collection involved in one spatial query and improve the efficiency of data retrieval.In data service, considering the requirement of data scheduling for virtual city visualization, we design and implement a model data engine based on spatial query, and relevant pre-reading and caching mechanisms are introduced to improve the data throughout. The model data engine runs as a local service, it connects the model database and the visualization client, receives data request from the client and loads data from the database and then returns the results to the client.Finally, we successfully integrate the model data engine to the Spatial Information System Platform which is developed by the laboratory independently. It provides solid data service for the visualization module and also works well for the application of large scenes.