In our virtual space teleconferencing system, in order to detect facial expressions and to reconstruct 3-D facial images, 3-D markers are attached to the face and traced visually. For the reconstruction of facial expressions, the tracing results are used to control the nodes of the wire frame model. In this paper, we explain the marker tracing system which consists of a CCD camera, a marker tracing device, and a personal computer. Markers are traced by this system at the sending site. Also we explain the facial image reconstructing system at the receiving site.
View full abstract