摘要
视频融合是视频监测控制相关应用领域关注的虚拟现实(VR)热点问题之一。根据实现虚实融合的维度不同,将相关研究方法分为4类:视频标签地图、视频图像拼接、视频叠加到三维场景、视频融合到三维场景。介绍了VR视频融合技术方面的相关工作,认为虚实融合技术正在快速发展中,虚实信息的可视关联对于人类认知和人工智能都已表现出显著的提升作用,未来将可能作为一种基础的地理信息资源来提供,有着重要的应用价值。
Video fusion is a hot topic in some virtual reality(VR)application areas such as video surveillance.According to dimension number,related works in virtualreality integration are surveyed in this paper in four categories:map with video tag,video image stitching,video overlapping on3D scenes,and the video fusion in3D scenes.Then our work in VR video fusion is presented in this paper,especially the new idea of video model and mixed reality rendering.The virtual reality technology is developing rapidly,and the visual correlation of the virtual and real environments has shown significant improvement on both human cognition and artificial intelligence.In the future,the video model may be provided as a fundamental geographic layer resource,which has significant values for various applications.
作者
周忠
孟明
周颐
ZHOU Zhong;MENG Ming;ZHOU Yi(Beihang University, Beijing 100191, China)
出处
《中兴通讯技术》
2017年第6期6-9,36,共5页
ZTE Technology Journal
基金
国家高技术研究发展("863")计划(2015AA016403)
国家自然科学基金(61572061
61472020)
