摘要
视频具有表达直观、信息丰富、动态实时等特点,将其与GIS有机集成形成地理视频(GeoVideo),可更加形象的表达地理空间。该文利用视频分割技术将视频分割为具有不同摄像机运动状态的视频片段,针对摄像机运动的视频片段,采用摄像机跟踪(Camera Tracking)技术求算视频帧对应的摄像机内外参数,并基于地理视频数据模型,采用KML/XML描述视频帧、摄像机位置、摄像机姿态等之间的时空映射关系。最后,设计了原型系统,在网络环境中聚合地理信息服务、视频流媒体服务与KML文档服务等,实现了视频与三维地理场景的交互与融合。实验表明,基于该地理视频数据模型,利用视频数据可实现对地理场景的增强表达,使用户能更形象的认知地理空间。
Video clip can provide rich information by the means of intuitive expression, real-time expression,and so on. GeoVidco, the video with geographic description, is a new approach to the geospatial representation. This paper designs a GeoVideo data model of interactive geospatial expression between video and 3D geographical scene. Video files were divided into video clips by different camera motions using video segmentation technique. The camera parameters of video frames were estimated through camera tracking technique. Then KMI. documents were used to describe mapping relationships among video frames, camera po- sitions and camera poses based on this GeoVideo data model. Finally, a web service-based prototype system was developed. These types of web services, i. e. 3D web mapping service, web video service and KML documents service, were ＇mashuped＇ using google earth API. Users can experience the interaction between 3D geographical scene and video, this overcomes the tradi- tional abstraction of 2D GIS and virtuality of 3D GIS, so people can more intuitive understand the geographic environment. Case study shows that the data model in this paper is feasible for augmenting aerial earth maps using videos, and this data naodel has application potentials in facility management, tourism promotion and other fields.
出处
《地理与地理信息科学》
CSCD
北大核心
2012年第5期6-9,F0003,共5页
Geography and Geo-Information Science
基金
国家科技支撑计划项目"视频GIS与突发公共事件的感知控制系统"(2012BAH35B02)
江苏省高校自然科学重大基础研究资助项目(10KJA420025)
