Vision Based Remote Control of Cellular

This paper describes the development and design of a vision-based remote controlled cellular robot. Cellular robots have numerous applications in industrial problems where simple inexpensive robots can be used to perform diierent tasks that involve covering a large working space. As a methodology, the robots are controlled based on the visual input from one or more cameras that monitor the working area. As a result, a robust control of the robot trajectory is achieved without depending on the camera calibration. The remote user simply speciies a target point in the image to indicate the robot nal position. We describe the complete system at various levels: the visual information processing , the robot characteristics and the closed loop control system design, including the stability analysis when the camera location is unknown. Results are presented and discussed. In our opinion, such a system may have a wide spectrum of applications in industrial robotics and may also serve as an educational testbed for advanced students in the elds of vision, robotics and control.