会员注册 会员登录 欢迎您,游客!您可以选择
最新公告:欢迎您访问同济大学软件学院图形图像研究中心(Smart3D Media Lab)网站。

您当前的位置:首页>>项目展示 项目展示

Smart_Web3D_BIM
图片展示
详细介绍

    本项目突破了轻量级Web3D-BIM关键技术(轻量化预处理、细粒度化预处理、渐进式传输、增量式网页级渲染处理等)成功地将大规模BIM场景放置在移动互联网上,用户无需下载安装任何插件,仅仅点击网页即可轻松在线浏览(甚至那些在单机用Revit等也很难打开和浏览的)大规模BIM场景。该技术使得智能城市的规划、设计、施工、审核、养护、竣工、运营等透过轻量级BIM可视化技术在移动互联网有机地“整合”起来了,实现了“Internet+”、“VR+”和“BIM+”的集成,其应用前景无限光明。

作为建筑业信息化的产物,BIM(Building Information Modeling)已成为智能构建的核心技术。结合大数据场景的冗余、语义、参数化和网络带宽、网页浏览器资源有限等特点,提出一套基于互联网+Web3D的地下空间BIM场景在线展示轻量化解决方案。


关键技术之一:轻量化预处理

    以IFC(Industry Foundation Classes)为数据输入格式,首先通过语义分析、几何重用去除和参数化表达等方式对BIM大数据进行轻量化预处理,从而大幅度地降低了BIM场景文件的数据量,最大程度地减少了BIM文件在互联网上传输的带宽消耗。



关键技术之二:细粒度化预处理

    首先,通过多视点投影分析将室内外场景数据分离,这样漫游外景时就无需加载内景,减少带宽与内存和渲染上的无谓消耗。其次,为方便进行可视化剔除,使用了稀疏体素化思想生成3D体素网格索引;设计了与视锥结合的兴趣区域FOI(Frustum of Interest) 管理策略,使得BIM场景的每一步漫游时仅仅动态加载可视增量,同时提出不可见减量场景,从而使得漫游时动态加载的数据量降为最低。最后,针对大规模室内场景进行了分块处理,并对其采用了数据文件打包,更为合理有效地使用互联网带宽,使得传输效率一步一步地提升,最终达到满足在线浏览与互动的目的。


关键技术之三:渐进式网页在线渲染

    考虑网页浏览器计算效率非常低,内存空间小等问题,我们采用了实例化渲染与渐进式渲染相结合的策略,边渲染边加载的策略,使得网页能够在线实时渲染大规模BIM场景,甚至能达到60FPS以上。总之,本项目成功地将在离线单机的Revit上很难渲染甚至很难打开的BIM文件在互联网上(网速正常时)可以很流畅地远程打开并随意浏览。


Project WebPage:

       http://koowii.net/tongyanproject/


Profile:

School of Civil & Building in Tongji University is the most famous research institution in the worldwide way. Smart3D Media Lab, as another leadership of lightweight Web3D key technology in Tongji University, aims at real time interactive visualization of Big BIM models over Internet. A hybrid team of Smart-Building and SmartWeb3D in Tongji developed a Pervasively Collaborative Smart3D BIM Visualization Platform on Mobile Internet for building planners, designers, constructors, investors, supervisors, end-users by accessing mobile Web Browsers anytime and anywhere, that will bring BIM with radical innovations by integrating “Internet+”, “BIM+” and “VR+”.


Project Summary

In this project, we present a Web3D-based lightweight solution for real time visualization of big BIM data scene, on considering its redundancy, semantics, parameterization but the limited resources of the network bandwidth and web browsers.

Firstly, taking industry foundation classes (IFC) as the input data format, we conduct a lightweight preprocess on raw BIM data via semantics analysis, reused object removal and geometry data parameterizing.

Secondly, we extract the space structure from the raw building for visibility culling and construct a scene index based on the sparse voxelization.

Thirdly, we integrate the above together in a progressive management strategy to update the scene data in real time.

The final system demonstrates: (a) with the semantics information, our method enables to significantly reduce the redundancies of the raw BIM big data; (b) the scene index supports data access and facilitates the indoor/outdoor visibility culling efficiently; (c) our management strategy does not only integrate our prototype system but also introduce a progressive management idea to improve the efficiency of resource consumption.

 

评论(0)

后面还有条评论,点击查看>>
图形图像研究中心版权所有
联系地址:上海市曹安公路4800号同济大学济事楼316室 电话:021-69583731