| The image as a feedback obtained by vision sensors is used by robot’s visual servoing to establish robot’s position closed-loop control and make the robot get a desired position.The robots identify grab and tracking control can be achieved on the basis of this.It contains many fields,such as machinery,electronics,image processing,sensor,information processing and communication and so on.Now,there are many problems in the domestic theory and practice,so the uncertainty of system model must be overcome in visual servo control.Then the control precision of robot can be improved.Visual servo system can be divided into monocular and binocular visual servo according to the number of cameras.The servo experiment tracking grab institutions of monocular and binocular was designed in this paper.The background of the paper is aimed at the needs of assembly line servo platform.The point movement servo control experiment platform of full closed-loop feedback was designed.The most outer ring of this system is the image control loop.The real-time image information of the goal is obtained by the CCD camera which is fixed on the end of the mechanical arm.Then the information is feedback to the industrial PC with PCI interface for image processing operations.The location information of the goals which is obtained from the image information is transmitted to the servo motion control system.The feature points of the image are converted to coordinate,and then the control commands are transmitted to the drive.The point movement of the mechanical arm is conducted under the control of drive.So the tracking control of the goal is completed and the goal of experiment is achieved.The overall analysis was conducted in view of the need of the paper,and then the overall design scheme of the system was determined.The key hardware of system was chosen and its software’s analysis,design and modeling were conducted through UML.The image processing and software processing and design of control subsystem are achieved on the basis of visual studio.Gaussian filter can hand noise well which was verified in image processing stage.The feature points of the target were detected,and then the coordinates were obtained through canny edge algorithm and Hough transform.The experiment control platform of servo system is GTS-400-PV(G)-PCI motion controller which is produced by Gugao Company.It can complete the point movement of the three dimensional coordinate system.Only the object tracking of the two-dimensional is studied in this paper.Its core is composed of FPGA which can realize high precision control.The servo experiment platform which is applied the robot vision technology is established on the basis of the research above.The point movement of the two dimensional planar manipulator was achieved,and the accuracy and effectiveness of the servo experiment was verified. |
PDF Full Text Request