基于低空全景视频的地理环境模型构建与可视化技术研究

项目名称： 基于低空全景视频的地理环境模型构建与可视化技术研究

项目编号： No.41201390

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 地理学

项目作者： 李科

作者单位： 中国人民解放军信息工程大学

项目金额： 25万元

中文摘要： 近年来，随着社会信息化进程的加速，各领域对地理信息的需求越来越高，特别是军队、公安、消防、交通等部门应对重大突发事件、自然灾害的处理时，要求测绘部门能够提供实时、动态、连续、大范围、可量测、二维正射与三维透视相统一的地理信息的支撑，这些要求已大大超越了传统基础地理空间4D产品所涵盖的范畴。 因此，本课题将以无人飞行器等低空平台获取的全景视频作为数据源，以视频技术、全景技术、地理信息技术结合为切入点，将视频序列影像中的每一帧影像经过自动化建模处理形成具有空间数学基础的全景图像；在视频时间轴上，建立视频片断与地理空间坐标框架间的坐标映射关系，在此基础上构建一个地理空间数据与全景视频融合、具有可量测、交互控制等功能的全景视频地理环境。在技术层面上，应用了最新的航空摄影技术、计算机图形图像技术、地理信息技术，实现了多学科集成；在科学层面上有助于虚拟地理环境理论、技术的发展，开拓新的研究方向。

中文关键词： 全景视频；数据融合；特征提取；视频分割；

英文摘要： With the fast development of information industry in the past few years, geographic information are used in more and more fields, especially when army, police, fire fighters, traffic department and so on are put their eyes on some big expected events and nature disasters. It is desirable for surveying and mapping department to supply geographic information support in both 2D and 3D forms, which must be real-time, dynamic, constant, large range, measureable. These requests have excess the category of traditional fundamentally spatial 4D production. Therefore, this task takes the panorama videos screened by unmanned aerial vehicles as data source, and do some research about video technology, panorama technology and geographic information technology. We take out every frame image form video serial images, which is modeling into a panorama image with math elements. In the time axis, we establish relationship between video snippet and geographic spatial coordinate frame. Based on these work, we can get a panorama video geographic environment, which can be measureable and interfaced with, combined by geographic spatial data and panorama video. In the technology lay, we use the newest aero photography technology, computer graphics and images technology and geographic information technology, and realize integration o

英文关键词： Panorama video；Data fusion；Feature extraction；Video Segmention；