Research Of Robot Arm Sorting System Based On Vision

With the continuous development of China's industrial production technology,the automation technology including the mechanical arm is more and more widely used in the assembly line,mainly used for the workpiece handling,grasping and other production behavior,the configuration of the mechanical arm is also more and more intelligent.In this paper,a grasping system based on visual manipulator is designed and verified by experiments.As a small grasping system,this system is of great significance to the application of machine vision grasping system in industrial production.It mainly consists of two modules.One is Open MV vision module,which is used to capture the moving three-dimensional object on the conveyor belt and carry out image processing to obtain the coordinates of the object's center point.The second is the robot arm module controlled by Arduino.The control board gets the position coordinates of objects through serial port communication,and then controls the work of steering gear to drive the robot arm to complete the grasping operation.When setting up the actual research environment,this topic focuses on the selection of software and hardware and analyzes the operation principle of software and hardware.As the basic work of the whole system,it analyzes the possibility of realizing the location information transmission through serial communication.Then,kinematics analysis and simulation of the robot arm are carried out,as well as the trajectory planning of the robot arm's grasping action.By analyzing the running speed of the conveyor belt and the position of the robot arm relative to the camera,the optimal grasping position of the object is calculated.At last,the coordinate position information of objects is obtained through three target recognition methods of template matching,color recognition and Arpril Tag,and the success rate of these three target recognition methods in grasping operation is analyzed respectively.