Research On The Manipulator Visual Servo System In Locating And Tracking

Intelligent robots is an important direction for today’s robotics research, therobot visual servo is a hot sub-topics in robot intelligent. This article does someresearch in manipulator visual servo system about motion control video imageprocessing and visual servo, then design motion controller and visual servocontroller to realize the object positioning and tracking.Firstly, the subject researched the manipulator motion control system, infeedback loop, the detection of the optical encoder signal based on state transition inorder to acquire the position and speed of robot arm. In the former channel to thecontrol system adopted the structure of double closed to design the control law. Thecontrol system indicators was obtained through the analysis of target motion andgeometric features, then simulated and debugged the control system in theenvironment of MATLAB RTW semi-physical simulation.Secondly, the project used Davinci DM6446image processing platform toprocessing the video signal. The video processing algorithms combined the methodof background subtraction and color threshold segmentation, when developingapplication program, the software used multithreading mechanism provided byLinux operating system to meet the real-time requirements of object tracking.Finally, this paper studied the camera calibration principle and determined therelationship between the camera coordinate and world coordinate, using the result ofextracted image features signal to design the visual servo controller, and thengenerated command of manipulator. Mismatch between the motion controlinstructions and system response, the article proposed three kinds of relativestrategies.