Research On Visual Calibration And Identification Accuracy Of Industrial Robots

Visual industrial robot is based on the addition of industrial robot visual module,visual module consists of two parts: hardware modules and software modules.Hardware modules include visual light source,industrial camera,computer three parts;software modules include image processing software and platform control software.Vision Industrial robots have "eyes" that recognize the material,follow the motion of the material,and grab the material quickly and accurately.Industrial robots with general visual accuracy cannot meet the needs of high-precision positioning and grasping.Therefore,it is of great importance to improve the visual accuracy of visual industrial robots in order to improve the accuracy of the grasping.In this paper,the circular PCB board as the research object,carry out visual industrial robot calibration accuracy and recognition accuracy.The main work of this thesis:1 The methods of improving the image quality,preprocessing the image and improving the edge extraction algorithm are used to study the image edge extraction accuracy.The results show that it can effectively improve the image edge extraction accuracy.2.The inner and outer circles of the PCB of the research object are respectively identified by the least square method and the Huff transform method,which effectively improves the image recognition accuracy.3.The use of checkerboard calibration plate to calibrate the image coordinate system,thus improving the visual image coordinates and industrial robot coordinates conversion accuracy.After the above research,the recognition accuracy of the image and the calibration accuracy of the system are improved,which improves the system accuracy of the visual industrial robot as a whole.The method and theory of this dissertation have a great guiding significance for the precision of the circular material,at the same time,it has some reference and reference significance for the precision of the material.