Large scale scene visualization is a challenge to Web3D virtual reality (VR) system. In the virtual urban simulation, apart from the buildings on the ground, the underground space visualization is also essential, because it includes many important components such as subway stations, car parks and underground utilities (e.g., pipelines and tunnels). This paper proposes a lightweight Web real-time rendering scheme for large scale underground space. Firstly, in order to organize the mass underground space data more efficiently, this paper proposes a lightweight scene ground structure. Then, this paper proposes an underground space scene management strategy based on pro-gressive area of interest (PAOI). In order to optimize the rendering system with respect to data loading according to needs, the components of the outdoor structure are bounded on their corresponding grid. Alternatively, the interior space is represented as a connected graph. Using the proposed techniques, the memory consumption is apparently cut off. Meanwhile, the display content significantly increases. To verify the effectiveness of the proposed algo-rithms, this paper implements the prototype system using Flash 3D. Without any plug-in installation, the user experi-ence is improved. The experiment shows that the proposed method has high performance and is suitable for large scale underground scene online rendering.%在移动互联网上显示大规模虚拟世界总是极具挑战性的课题。在虚拟城市模拟中，除地面建筑外，地下空间可视化也是必需的，因为它包括很多重要的组件，例如地铁站、停车场以及其他地下设施（如管线和管道）。针对大规模地下空间场景，提出了一套适合在网页级实时绘制的轻量级解决方案。首先，对地下空间种类庞杂而又弱关联的数据模型进行了梳理与组织，建立了一套适用于Web3D显示的轻量级场景结构；然后，提出了地下空间的渐进式兴趣区域网格化场景管理策略（progressive area of interest，PAOI）。为了实现按需加载思路，在地下空间室外场景，算法将地下室外构筑物与其所在网格绑定；在地下空间室内场景，将室内场景通过连通面结构进行剖分组织，建立索引。通过以上技术，从整体上大大减少了内存开销，明显增加了显示内容。使用了渲染计算能力相对薄弱的Flash3D实现了所提出的算法。实验结果证明了该方法的有效性，且适用于大规模地下空间场景的互动展示。
展开▼
