3D motion and shape representations in visual servo control

The study of visual navigation problems requires the integration of visual processes with motor control. Most essential in approaching this integration is the study of appropriate spatiotemporal representations that the system computes from the imagery and that serve as interfaces to all motor activities. Since representations resulting from exact metric reconstruction of the environment have turned out to be very hard to obtain in real time, the authors argue for the necessity of representations that can be computed easily, reliably, and in real rime and that recover only the information about the 3D world that is really needed to solve the navigational problems at hand. In this paper; the authors introduce a number of such representations capturing aspects of 3D motion and scene structure that are used to solve navigational problems implemented in visual servo systems.