Robotic vision can increase high-tech ingredient and improve the detection precision of robot. It is an important issue in industrial automation and has drawn particular attentions.Aiming at trajectory tracking with robotic vision, this dissertation gives some conclusions on the research of the position-based system. To track a curve, the control is decomposed into two subfunctions in the small window created according to the theory of pattern recognition. One is the velocity control which controls the center of a CCD camera to move along the tangential direction of the curve. The other is the position control which controls the center of the CCD camera to move in the normal direction approaching to the curve.And it also presents some methods of image process, including median filter and gray binary, gives matlab procedure on camera calibration, and analyses the robotic visual servo system. It establishes the basement for the further research on 2D-line tracking system.
|