Research On Differentiated Mobile Phone Middle Frame Grabbing Technology Based On Visual Guidance

With the increasing automation of mobile phone assembly,the production line uses industrial robots for loading and unloading operations in gradually.At present,industrial robots are mainly rely on teaching online operations,their movement mode and path are set by teaching in advance,and then repeat point-to-point operation.This kind of operation mode has low flexibily and high professional requirements for teaching and offline programming.When positioning and grabbing different materials,the adjusting time takes more,which affects the efficiency of production switching.In this paper,in order to solve the problem of poor adaptability of loading robots to differentiated products in mobile phone automatic assembly lines,it was difficult to adjust the positioning and grasping of the middle frames of different models of mobile phones quickly.Therefore,this paper proposes a highly adaptive grasping and positioning method for visionguided cartesian robot.The main research contents were shown as follows:Firstly,this paper introduces the function requirements and technical indexes of the robot loading and unloading system and adopted the overall scheme combining of "offline visual analysis grabbing location" and "online visual positioning grabbing",and designed the workflow of the hardware and software architecture of the system in detail.Secondly,offline visual positioning was used to find the relative pose of the robot's grabbing of the middle frame of the mobile phone,focusing on the problem of the hole in the middle frame of the mobile phone affecting the stability of the grabbing.Specifically,in order to ensure the integrity of the outer edge contour and hole edge features in moboile frame,median filtering and morphological closed operation were selected according to different feature requirements.After analyzing and comparing several edge detection operators,Canny operator was selected to extract the outer edge and hole edge features of the middle frame,and the minimum circumscribed rectangle algorithm and the minimum circumscribed circle algorithm were used to obtain the vector information and hole feature information of the middle frame of the mobile phone.In addition,based on the hierarchical traversal search strategy of full but not complete binary tree,this paper proposes a grasping and positioning algorithm for the robot end effector to avoid grasping and locating holes in the middle frame,and determines the relative position and pose of the robot grasping of the middle frame of mobile phone,which lays the foundation for the follow-up online visual positioning.Thirdly,online visual positioning was used to guide the cartesian robot to recognize and grasp the target in this paper.Specifically,based on Hausdorff distance matching algorithm,this paper realizes the online recognition and positioning of mobile phone middle frame and the acquisition of pose information.In addition,the nine-point calibration method was used to randomly collect the image coordinates of the target points on the nine groups of target materials and the corresponding robot coordinates,and the calibration conversion matrix was solved by the least squares method to realize the uniform of coordinate positions.Moreover,this paper designed a calculation method for the compensation of the motion position of the robot end effector,and completed the robot grasping motion path planning to realize the accurate grasp of the mobile phone frame.Finally,based on the development environment of Visual Stutio 2017,this paper develops the function modules of visual positioning system,introduces the operation methods of the functional modules such as image acquisition,visual calibration,template creation,and communication were introduced in detail,and realizes the integrated control communication between the vision system and robot.The four sets of experiments were used to verify the vision-guided grasping and positioning scheme proposed in this paper has good adaptability and can meet the accuracy requirements of robot feeding.