An Auto-Stereoscopic VRML Viewer for 3D Data on the World Wide Web

This paper describes an auto-stereoscopic VRML viewer for a stereoscopic display without any glasses called “Rear Cross Lenticular 3D display (RCL3D display).” By presenting a virtual environment directly on the 3D display, people feel the depth of environment just as the physical model. In the “net-surfing,” this viewer automatically generates a pair of stereoscopic images from VRML2.0 data on WWW sites and composes them into one image for the 3D display. Operations such as the walk-through and interaction in a virtual environment are provided. In VRML2.0 data, information to generate the stereoscopic images, such as a base line length and a convergence angle, is not represented. Thus, this viewer provides a way to dynamically adjust such parameters by the operator in order to get natural depth feeling.