A new fuzzy visual servoing with application to robot manipulator

Many stereo vision algorithms require that images to be very similar in appearance and that the distance between the two cameras should be small enough. Normal visual servoing of a robot manipulator needs both image information and joint velocities. In this paper we make two modifications to overcome these problems: 1) a new simple stereo vision model is proposed, 2) a fuzzy controller based only on visual information is implemented to a robot manipulator. We successfully apply our new algorithms to real-time visual servoing of an industrial robot. The stereo model proposed in this paper can also be extended to other applications, such as mobile robots, 3D reconstruction, etc.