Design And Research Of Steel Structure Workpiece Recognition And Positioning System Based On Machine Vision

At present,prefabricated steel structure buildings account for the highest proportion and the greatest demand in the future.As the main component of assembled steel structure building,the demand of output and quality is increasing year by year.Intelligent data processing technology and machine vision technology is the hotspot upgrade traditional manufacturing technology,the intelligent data processing technology can ensure the whole production line of data demand,to maintain the integrity of the whole and the accuracy of the production,and machine vision technology can improve the recognition accuracy and efficiency at the same time,save manpower cost.Based on the technological background of the transformation of steel structure production mode,a steel structure workpiece recognition and positioning system based on machine vision is designed in this paper.The main research contents are as follows:First,according to the production mode and structural characteristics of building steel structure,determine the functional requirements of the system,and design the required modules and functions of the whole system.A positioning system of steel structure recognition machine based on machine vision is designed by integrating data processing,database technology and machine vision detection technology.Secondly,MySQL database tool was used to complete the creation of steel structure workpiece database,and Python programming language was used to write the steel structure 3D data XML file parsing program,successfully obtained the steel structure workpiece characteristic information,positioning information and the workpiece contour information taken by the simulation camera.The final positioning information of the workpiece is converted into attitude matrix,Euler and quaternion and stored in the database.At the same time,the automatic recognition and positioning system based on machine vision is developed by combining database data and actual workpiece.Python was used to connect daheng industrial camera to complete image collection,image processing flow was designed for the actual workpiece image,and contour extraction and feature information extraction of the actual workpiece were realized.Using Python programming language to write the workpiece recognition and location algorithm,complete the recognition and location of the workpiece to be welded.On this basis,an experimental system of steel structure workpiece identification and positioning based on machine vision is established.The system can correctly identify the experimental work piece,obtain the information of the work piece,determine the position and pose of the work piece and transmit it to the industrial robot for grasping,and obtain the final position and pose of the work piece in the steel structure.The research content of this paper can provide reference for the application of machine vision in the identification and positioning of steel structure workpiece.