面向虚拟城市的三维模型管理技术研究

发布时间：2018-01-15 11:40

本文关键词：面向虚拟城市的三维模型管理技术研究　出处：《浙江大学》2015年硕士论文　论文类型：学位论文

【摘要】：三维城市的建设和应用,已经成为信息技术发展的必然趋势,虚拟城市环境的模拟也是“数字城市”建设的重要组成部分。虚拟城市的构建需要大量的模拟城市地物景观的三维模型数据,模型数据的组织和管理是最为基础同时也是比较关键的技术之一,直接影响着模型数据可视化的实时性。本文将着重研究城市三维模型数据的组织和存储模式以及如何构建高效的模型数据服务。在模型数据对象的组织方面,传统的方法是以一个模型文件为单元构建模型数据对象,本文针对三维模型数据的数据量大且数据密度高的特点以及面向虚拟城市可视化应用的调度需求,构建了相比文件而言粒度更小的模型数据对象单元并且采用适用于可视化调度的数据存储结构。一方面,增加了调度的灵活性,另一方面也避免了数据加载时格式化的问题。在模型数据的管理方面,通常有两种方式即基于空间数据库的管理方式和基于文件系统的管理方式。本文分别设计和实现了这两种数据管理方式。其中,对于基于空间数据库的方式,结合CityGML对城市模型数据对象的描述和数据调度的需求,设计实现了相应的空间数据库模式,模型数据对象结构更加合理,可以满足局部小区域数据的组织、管理和调度工作；而对于基于文件系统的方式,本文将其扩展到分布式文件系统上以适应未来数据不断增长的需求,并且针对数据调度的局部性特征,设计了基于网格划分的模型数据组织方式,并构建了基于网格划分的空间索引体系,减少了一次空间数据查询涉及的数据集,提高了数据检索的效率。在数据服务方面,针对虚拟城市可视化应用的数据调度需求,设计实现了基于空间查询的模型数据引擎,并引入了相关的预读和缓存策略,以提高数据的吞吐量。模型数据引擎以客户端的本地服务形式运行,将可视化客户端和模型库连接起来,接收客户端的数据请求并访问模型库获取数据后返回给客户端。最后,本文实现了模型数据引擎和实验室开发的空间信息系统平台的集成,为空间信息系统平台的可视化模块提供稳定的数据服务,在大场景的可视化应用中也能有较好的性能。
[Abstract]:The construction and application of three-dimensional city has become the inevitable trend of the development of information technology. The simulation of virtual city environment is also an important part of the construction of "digital city". The construction of virtual city needs a lot of 3D model data to simulate the urban landscape. The organization and management of model data is one of the most basic and key technologies. This paper will focus on the organization and storage mode of urban 3D model data and how to construct efficient model data service. The traditional method is to construct model data object with a model file as a unit. This paper aims at the characteristics of large data volume and high data density of 3D model data and the scheduling requirements for visual application of virtual city. The model data object unit with smaller granularity than the file is constructed and the data storage structure suitable for visual scheduling is adopted. On the one hand the flexibility of scheduling is increased. On the other hand, it also avoids the problem of data formatting when loading. In the management of model data. There are usually two ways: spatial database based management and file system-based management. This paper designs and implements these two data management methods respectively. According to the demand of CityGML for the description of urban model data object and data scheduling, the corresponding spatial database model is designed and implemented, and the model data object structure is more reasonable. Can meet the local small area data organization, management and scheduling work; For the file system-based approach, this paper extends it to the distributed file system to meet the growing demand of future data, and the local characteristics of data scheduling. The model data organization method based on grid division is designed, and the spatial index system based on grid partition is constructed, which reduces the data set involved in a spatial data query. The efficiency of data retrieval is improved. In the aspect of data service, a model data engine based on spatial query is designed and implemented to meet the data scheduling requirements of virtual city visualization applications. In order to improve the throughput of the data, the model data engine runs as the local service of the client and connects the visual client and the model base. After receiving the data request from the client and accessing the model library to get the data, it is returned to the client. Finally, the integration of the model data engine and the spatial information system platform developed by the laboratory is realized. It provides a stable data service for the visualization module of the spatial information system platform, and can also have good performance in the visualization application of large scene.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.9

【共引文献】