Design Of The Motion Control And Grasping System For Robot Arm Based On Vision

In recent years,with the rapid development of artificial intelligence technology,the application value of robotic arm has become more and more important.However,the traditional robotic arm can only grasp the target according to the planned path and can not get the outside information,so in the case of unknown or changed surroundings,it is necessary to readjust the manipulator,otherwise it will be difficult to complete the grasping task.In order to strengthen the flexibility of the manipulator control system and make it have the ability to perceive the surroundings,the robotic arm control technology based on vision has gradually become the mainstream of the development of robot industry,and it lays the foundation for solving the problem of robotic arm grasping in unknown environment.Aiming at the problem of manipulator control in unknown environment,based on the Kinect vision sensor and using the robot operating system(ROS),a visual positioning and target grabbing system of six-DOF manipulator is designed,which is referred to as the robotic arm visual grasping system.Firstly,according to the actual demand of the robotic arm vision grasping system,the system design scheme is proposed,and the selection and analysis of the Visual positioning module and the motion control module are made.Secondly,the related technologies of object recognition and location are studied in detail by using color segmentation,including Kinect camera calibration,color segmentation based on HSI space,feature extraction and crawl point location,etc.Thirdly,using d-h parameter method to complete the model of manipulator,the inverse kinematics of the manipulator is solved and analyzed,and then the trajectory of the manipulator is programmed by the linear interpolation and circular interpolation movement method of the Cartesian space,and the Rviz tool is used to simulate the trajectory planning algorithm of the manipulator in the ROS environment,Thus laying the foundation for the motion trajectory planning of solid manipulator.Then,the software modules such as visual positioning and manipulator motion control are designed;finally,the visual grasping function of the manipulator is realized.This system takes the cylindrical box as the positioning target,carries on the multiple vision localization and the picker grasping experiment under the laboratory environment,the test data shows that the precision of the robotic arm vision grasping system satisfies the requirement of the six-DOF manipulator to the target object visually grasping.At the same time,the grasping error is also within the permissible range of the manipulator's end actuator grasping the target object.