中文摘要：

针对无人机（UAV）影像等多源影像构建全球虚拟场景的数据组织和调度问题，研究了无人机影像的预处理方法，讨论了适于可视化需要的球体简化投影方法以及全球虚拟场景的结构，并在此基础上提出了支持无人机影像的全球场景多线性四叉树的组织方式，介绍了无人机影像等多源影像跨源影像衔接与调度的关键技术。用5．12汶川大地震中获取的无人机影像数据进行的实验证明，采用文中提出的数据处理与组织管理方法，能够减少数据存储的冗余性，提高系统运行的效率，方便数据的管理与更新，高效地提供灾情信息可视化服务。

英文摘要：

Aiming at the data organization and data dispatch problems in building a global virtual scene with multi-source images including tmmanned aerial vehicle （UAV） images, this paper explores the UAV image pre-processing, the simplified sphere projection suitable for visualization, and the structure of the global virtual scene. Based on this, an UAV imagesupporting structure of the model for multi-quadtree organization in the global scene is proposed, and the key techniques for inter-source linking and dispatching of the data of multi-source images including UAV images are given. The result of the experiment performed using the UAV images acquired in the 5.12 Wenchuan earthquake as the data source shows that the methods presented in the paper can reduce the data redundancy, promote the system running efficiency, make data management and updating convenient, and effectively provide visualization service in disaster response and management.