基于视频模型的虚拟现实视频融合系统  

Virtual-reality Video Fusion System Based on Video Model

作　　者：周颐[1] 孟明 吴威[1] 周忠[1] Zhou Yi;Meng Ming;Wu Wei;Zhou Zhong(State key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100083, China)

机构地区：[1]北京航空航天大学虚拟现实技术与系统国家重点实验室,北京100083

出　　处：《系统仿真学报》2018年第7期2550-2557,共8页Journal of System Simulation

基　　金：国家863计划(2015AA016403);国家自然科学基金(61572061;61472020;61502020);中国博士后科学基金(2013M540039)

摘　　要：虚实融合技术以减少虚拟对象和真实视频图片之间的视觉差异为目标,力图实现视觉感官上的无缝结合。多视频的虚实融合技术普遍存在画面畸变与虚实对准代价大的问题。针对该问题,提出并设计了一种基于视频模型的虚拟现实视频融合系统。介绍了视频的注册与模型建立原理,阐述了系统中视频与场景融合绘制方法。对系统的总体结构进行了介绍。在实验阶段,测试了本系统的性能,并与国际上的虚实融合方法进行了对比分析。表明该方法在融合绘制效果和性能上更优。The aim of virtual reality integration is to reduce the visual difference of virtual object and real video image. However, it still has several problems, including heavy imagery distortion when the user deviates from the source viewport, and huge modeling cost of virtual scene. For these issues, this paper proposes a novel virtual-reality video fusion system based on video model. We firstly introduce how we register and build video model with interactive operations, then present the fusion method of video and virtual scenes. On the design and implementation stage, we introduce the architecture of our system. Finally, we test our system＇s render performance and compare it with two state-of-the-art virtual reality integration methods, which demonstrates that our fusion method is better than other two methods.

关 键 词：视频投影 视频模型 纹理贴图 图片建模 虚实融合技术 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]