摘要
综述了基于图像的虚拟现实环境的建模技术 ,并给出了各种模型表示的理论依据和建模过程的定义。把环境表示分为四类 :全景图、光场数据库、深度图像和多投影中心图。全景图的生成包括基于图像相关性和基于图像整合变换的算法。光场数据库的分类主要依据全光函数。同时论述了在各种表示下的采样和建模技术。
We survey the techniques for image based modeling on VR(virtual reality). We present the theory basics of the representations and the definition of modeling procedure. We divide the virtual reality representations into four types: panoramic image, light field, depth image, multiple centers of project images. The generation of panoramic image includes correlation based techniques such as SSD and image registration techniques. The classification of light field mainly depends on plenoptic function. We discuss how to sample and model VR. At last, we discuss the major problems that will be overcome in the future
出处
《高技术通讯》
EI
CAS
CSCD
2002年第11期107-110,共4页
Chinese High Technology Letters
基金
国家自然科学基金 (60 0 75 0 10 )
国防预研基金 (0 0J4.5 .6HT0 10 9)资助项目
