Video texture cues enhance stereoscopic depth perception in a virtual reality-based, telerobotic interface

摘要

A major challenge in our telerobotic research is helping operators understand complex spatial relationships at the remote robot work site. Many researchers have reported that the technology of field-sequential displays, such as stereoscopic goggles, can be useful when depth perception is important. In applying this technology to our work, we met and explored a major limitation to field-sequential displays: the effect of target orientation. Stereoscopic depth cues do not define the depth of large, homogeneous, horizontal objects. The variance of operator's depth estimates for the same target increases as the orientation of the target's major edge is rotated from a vertical position to a horizontal position. However, drawing an appropriate texture on the surface of the feature completely overcomes the orientation factor. Further experiments explore some critical features of the texture, including texture coarseness and orientation. A final experiment shows the benefit of etching textures directly onto the surface of the graphical model compared to superimposing a texture on a window looking into a 3D model.