三维国土管理云平台多尺度模型和渐进式传输关键技术研究

发布时间：2018-06-22 07:11

本文选题：三维国土云平台 + 多尺度模型　；参考：《电子科技大学》2017年硕士论文

【摘要】：三维国土管理云平台是国土信息化发展的必然趋势,大量国土数据的网络传输和实时渲染是平台面临的一个巨大挑战。多尺度模型和渐进式传输技术是解决大规模国土数据的网络传输和实时渲染问题的关键技术。论文实现了大量国土数据的多尺度模型构建、渐进式传输、三维可视化,为三维国土管理云平台的研发奠定了坚实的技术基础。同时采用数据预取、数据移除等技术,提高多尺度模型和渐进式传输原型系统的性能。本论文的主要研究工作如下:(1)根据Visvalingam-Whyatt算法本身的不足和论文使用矢量数据特点,提出了基于单调曲线的改进的Visvalingam-Whyatt算法,有效地解决了构建矢量数据多尺度数据时出现的自相交、占用内存大等问题。(2)借鉴多细节层次的多尺度模型,提出基于空间格网的多细节层次的多尺度模型,进行矢量数据的多尺度组织,解决了大量矢量数据的多尺度存储、渐进式传输、快速检索等问题。(3)讨论了渐进式传输中的关键技术:多线程技术、数据移除、数据预取。矢量数据的渐进式传输框架包含多个重要的任务,每个任务之间相互独立,通过多个线程协作来完成;数据量大,占用客户端内存大,通过数据移除算法来移除暂时不会加载的数据;为了避免数据加载缓慢,采取数据预取技术。(4)研究影像数据的预处理方法,包括波段提取、波段融合、影像拉伸、影像镶嵌、影像裁剪等;使用了基于行分割的影像金字塔算法,解决了影像切片缓慢、内存占用大等问题。(5)最后详细地论述了多尺度模型和渐进式传输原型系统,采用C/S架构,实现了国土数据的渐进式接收、重建、三维可视化;并包括国土数据多尺度构建工具,用以生成矢量和影像数据的多尺度数据。并给出了原型系统的性能评价方法,对系统的主要性能进行了对比,证明了系统的实用性。本论文研究并改进了多尺度模型和渐进式传输方法,开发出了原型系统,实验结果表明大量国土数据的网络传输性能较好,能够满足实际应用需要。
[Abstract]:The 3D land management cloud platform is the inevitable trend of the development of land information, and the network transmission and real-time rendering of a large amount of land data is a great challenge to the platform. Multi-scale model and progressive transmission technology are the key technologies to solve the problem of large-scale land data transmission and real-time rendering. In this paper, a large number of land data multi-scale model construction, progressive transmission, three-dimensional visualization, for the development of three-dimensional land management cloud platform laid a solid technical foundation. At the same time, data prefetching and data removal techniques are used to improve the performance of multi-scale model and progressive transmission prototype system. The main work of this thesis is as follows: (1) according to the shortcomings of Visvalingam-Whyatt algorithm and the characteristics of vector data, an improved Visvalingam-Whyatt algorithm based on monotone curve is proposed, which effectively solves the self-intersection in constructing multi-scale data of vector data. (2) drawing lessons from the multi-scale model of multi-level of detail, a multi-scale model based on spatial grid is proposed to organize the vector data in multi-scale, which solves the problem of multi-scale storage of a large number of vector data. (3) the key technologies of progressive transmission are discussed: multithreading, data removal and data prefetching. The progressive transmission framework of vector data contains many important tasks, each task is independent of each other and is accomplished by multi-thread cooperation, which has a large amount of data and takes up a large amount of memory on the client side. In order to avoid slow data loading, data prefetching technology is adopted. (4) the pretreatment methods of image data are studied, including band extraction, band fusion, image stretching, image mosaic, etc. Image clipping and so on; using the image pyramid algorithm based on line segmentation to solve the problems of slow slicing and large memory footprint. (5) finally, the multi-scale model and progressive transmission prototype system are discussed in detail, using C / S architecture, It realizes the gradual receiving, reconstruction and 3D visualization of land data, and includes the multi-scale construction tool of land data, which is used to generate multi-scale data of vector and image data. The performance evaluation method of the prototype system is given, the main performance of the system is compared, and the practicability of the system is proved. In this paper, the multi-scale model and progressive transmission method are studied and improved, and a prototype system is developed. The experimental results show that the network transmission performance of a large amount of land data is good and can meet the needs of practical applications.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P208

【参考文献】