Research On Positioning Method Of Angle-Weld Of Stainless Steel Based On Line-Structured Light

Welding is a important part of the industrial manufacturing process and one of the symbols of the national machinery manufacturing and technological development level.In order to improve productivity and production quality,welding robots began to replace manual welding.Weld positioning is a key technology for robotic automated welding,in which structured-light-based method has the advantages of high precision,fast speed and non-contact.According to actual research,it is found that the current vision-based seam positioning system has the disadvantages of low and insufficient applicability.Especially when measuring stainless steel welds,the specular reflection effect is serious,due to the high smoothness of the surface of the welding material.This makes the weld feature extraction and positioning accuracy insufficient,and the welding effects unstable.This paper studies the positioning method of stainless steel right-angle welds based on structured light vision,and designs and develops the corresponding structured light vision positioning system for welds.In this thesis,the status of structured light vision measurement technology and weld image processing technology at home and abroad are comprehensively studied.The device of stainless steel seam positioning system based on structured light vision is designed and the selection of key equipment is also completed.Based on the principle of optical triangulation,the structured light measurement method is designed,and the structured light vision system is constructed.The camera parameters are solved using the Zhang's calibration method,the calibration accuracy was doubled by reducing the interference of highlight area and secondary optimization after correction.A fast and effective calibration method of structured light plane was used to calculate the equation of light plane,and the calibration was completed.According to the parameters obtained from the calibration,the hand-eye relation was calibrated by the Eye in Hand method,and the hand-eye relation matrix is obtained.The structured light image features of stainless steel right-angle welds are analyzed.Bilateral filtering is used for noise suppression,and the gray histogram threshold method is used to extract the structured light bar.On this basis,the method of optimizing the shape of the structured strip and removing the reflection noise is proposed,then the preprocessing process is completed.The fast ZS algorithm was used to extract the skeleton of the ray strip,and the judgment algorithm was proposed to achieve the stable Hough center line fitting effect.The prior knowledge was used to identify and locate the seam feature points,and experimental results show that the average precision is 0.22 mm.The positioning method for stainless steel right-angle seam is stable and effective,which can meets the actual welding requirements.In order to realize the integration function of algorithm,the software of structured light vision system is developed based on Qt5 and Open CV2.