针对管道全位置机器人自动焊,提出了一种基于单目主被动视觉结合的熔化极气体保护焊(gasmetal arc welding,GMAW)焊接偏差测定方法。设计出一种调光玻璃,使采集到的同一帧焊接图像中包含了激光条纹和焊丝尖端等信息。首先,根据焊接图像噪声类型呈高斯分布这一特性,针对性的采用LoG算子分别对电弧区域和激光条纹区域图像进行滤波处理；然后,设计图像处理算法分别提取出焊丝和焊缝坡口中心位置坐标；最后,计算出两坐标点在Y轴方向的偏差值。试验证明,此方法能在填充焊焊接图像中实时测定出焊接偏差量,焊接误差可以控制在02mm以内,可为实现管道焊接机器人自动焊缝跟踪控制提供可靠依据。
Aiming at all position automatic welding of pipeline by robot,a welding deviation measurement method of gas metal arc welding(GMAW) based on monocular active and passive vision is proposed. A kind of dimming glass is designed,so that the same welding image contains the information of laser stripe and welding wire tip. Firstly,according to the characteristic that the noise type of welding image is Gaussian distribution,the LoG operator is used to filter the images of arc area and laser stripe area respectively；then,the image processing algorithm is designed to extract the coordinates of welding wire and weld groove center；finally,the deviation value of two coordinate points in Y axis direction is calculated. The experiment shows that this method can measure the welding deviation in real time in the welding image of the filling welding,and the welding error can be controlled within 0.2mm,which can provide a reliable basis for realizing the automatic seam tracking control of the pipeline welding robot.
KE Xi-lin, WANG Zhong-ren, LIU Hai-sheng, WANG Xiao-gang. Welding deviation detection method based on monocular active passive vision[J]. LASER & INFRARED,2021,51(11):1519~1525