Research On Visual Grasping Method Of Material Robot In Electronic Line

Robots are widely used in all aspects of industrial production due to their high efficiency and high precision.In the electronic automated production line,due to factors such as the shape of the material itself and the disorderly placement of the material,it is difficult to achieve automatic robot feeding.Manual feeding is the main method now.The loading robot is an indispensable robot on an intelligent production line.It plays an important role in stabilizing production cycle and improving production efficiency.Therefore,it is necessary to realize an automatic loading robot.This article takes the automatic assembly of the mobile phone camera and flash on the intelligent electronic production line as the background,and takes the mobile phone shell as the object automatically picked by the loading robot.It focuses on the visual positioning method and pose estimation method of the mobile phone shell.The main research content and results of this article as follows:Based on the research on the current development status and basic framework of robot gripping,combined with the characteristics of the mobile phone shell automatic feeding robot studied in this paper,a set of feasible feeding schemes is designed.The main components of the feeding robot are discussed,the camera calibration method and the hand-eye calibration method in the robot hand-eye system are studied,and related experiments are carried out,so that the parameters of the hand-eye system are accurately obtained,which is very important for the subsequent research.Aiming at the visual positioning of the mobile phone shell,due to the poor performance of the method based on the rectangular detection on the mobile phone shell with rounded corners and the inapplicability of the current general target detection method in the situation described in this article,an image-based geometry is proposed A mobile phone case detection and positioning method with constant moment,which calculates the minimum area circumscribed rectangle of the detected target to locate the center and vertex of the mobile phone case.Experiments show that this method can achieve a recall rate of 98.3% and a positioning error of 0.5 Within pixels.Aiming at the posture estimation of the mobile phone shell,a posture estimation method based on monocular vision is studied and implemented.This method uses the corresponding relationship between the feature points obtained by visual positioning and its object coordinate system to construct a PNP problem.It focuses on the use of the center of the mounting hole of the mobile phone housing camera to construct a reference vector,thereby obtaining the pixel coordinates of the feature points and the three-dimensional space coordinates.Correspondence,this method does not require a three-dimensional model of the object and only requires a monocular to complete the pose estimation of the target.The automatic feeding scheme for mobile phone shell robots of electronic production lines designed in this paper has high target positioning accuracy and practical pose estimation methods,and has high value for the application of feeding robots.