Virtual Environment Interfaces to Physical Environments: A Tool for Cooperative Research in Experimental Robotics

其目标是开发一种计算机系统，以支持地理上分散的研究人员在机器人领域进行合作实验研究。该系统在智能规划者的控制下使用视觉技术来获取实验所在物理环境的视觉模型(结合了3-D数据和纹理贴图)。它提供了在研究人员的图形工作站上沿着个别选定的视点进行实验的实时图形渲染。因此，每个研究人员都能够使用他/她自己的可控制虚拟相机来观察实验。最初的探索将集中在设想的系统的一个主要组件上，即自动构建一个实验机器人场地的3-D模型。这将被用来分析在设想的系统的其他组件中也将遇到的基本问题，特别是不确定性对用户与远程实验场之间交互的影响。这是宾夕法尼亚大学GRASH实验室的R.Bajcsy和斯坦福大学的J.C.Latombe共同努力的成果。

The goal is to develop a computer system to support geographically separated researchers performing cooperative experimental research in robotics. This system uses vision techniques under the control of smart planners to acquire a visual model (combining 3-D data and texture maps) of the physical environment where the experiment takes place. It provides real-time graphic renderings of the experiments on the researchers' graphic workstations along individually selected viewpoints. Hence, each researcher is able to OBSERVE the experiments using his/her own controllable virtual camera. The initial exploration will focus on one major component of the envisioned system, the automatic construction of a 3-D model of an experimental robotic site. This will be used to analyze basic issues that will also be encountered in the other components of the envisioned system, in particular the effect of uncertainty on the interaction between a user and a remote experimental site. This is a collaborative effort between R. Bajcsy of the GRASP laboratory at the University of Pennsylvania and J. C. Latombe at Stanford University.