首页 | 本学科首页   官方微博 | 高级检索  
     检索      

海量三维地形数据集的流式处理
引用本文:陈斌,董士海,方裕.海量三维地形数据集的流式处理[J].地理与地理信息科学,2005,21(2):20-23.
作者姓名:陈斌  董士海  方裕
作者单位:北京大学遥感与地理信息系统研究所,北京,100871;北京大学信息科学技术学院,北京,100871
基金项目:国家高技术研究发展计划(863计划)
摘    要:该文提出基于网络实时交互的海量三维地形数据的流式处理方案。通过一个输出恒定的动态多分辨率简化算法,对三维地形数据进行裁剪简化,实现与总数据量和视点位置无关的恒定数据量输出;在算法输出的非均匀高程矩阵上,进一步采用三维地形数据流式编码解码算法,将高程矩阵进行离散余弦变换到频率域,并根据频谱分析结果引入累进式传输编码解码,结合服务质量控制,根据客户端和网络的性能状况进行自适应调节。实验表明,流式处理方案可以根据运行平台性能提供不同质量的场景复原,达到基于网络的实时交互性能。

关 键 词:三维地形  流媒体  GIS  服务质量控制
文章编号:1672-0504(2005)02-0020-04
修稿时间:2004年10月27

Stream Processing for Massive 3D Terrain Dataset
CHEN Bin,DONG Shi-hai,FANG Yu.Stream Processing for Massive 3D Terrain Dataset[J].Geography and Geo-Information Science,2005,21(2):20-23.
Authors:CHEN Bin  DONG Shi-hai  FANG Yu
Abstract:Enlightened by stream medium technology,this paper proposes a method handling 3D terrain data in streaming way for real-time interaction based on network.At the beginning,a stable output depending on dynamic multi-resolution simplification algorithm is applied to get a series of fix-sized non-uniform elevation matrices.A streaming code based on DCT for terrain data is then introduced.The non-uniform elevation matrix is transformed by the code to frequency domain and results in a coefficient matrix.Then the result coefficient matrix is coded in a progressive format.Finally,together with QoS control,which is based on total system performance such as client CPU/rending capability and network traffic factors,the real-time interaction to massive terrain data through network is achieved.This stream processing method for terrain data also enriches traditional terrain interaction system by bringing features such as live or recorded interaction broadcasting,live interaction recording and editing.A sample program demonstrates different decode quantity of terrain scene for corresponding total system performance to achieve real-time interaction through network.
Keywords:GIS
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号