Research On Weld Recognition And Location Of Crane Main Beam Based On Machine Vision

Crane is one of the most important equipment in heavy industrial production.In the production process of crane,the welding process of inner seam of crane main beam has always been a difficult point in its production operation.Due to the narrow welding space of the inner seam of the main beam of the crane,manual welding is difficult,labor-intensive,and the quality cannot be guaranteed.The use of teaching robot welding also has problems such as difficult teaching operations and low production efficiency.Therefore,this subject uses machine vision technology to carry out research on the identification and positioning of the welding seam of the crane girder,which has great practical significance for improving the intelligence and flexibility of the welding production of the crane girder.The main work contents are as follows.(1)Research on the calibration method of machine vision system.A singlecamera dual-station hardware platform composed of a FNAUC industrial welding robot and a CCD camera was built.In order to solve the mapping relationship between the point of the crane girder weld in three-dimensional space and its pixel point on the two-dimensional image,a monocular camera imaging model was established,and the internal and external parameters of the camera were obtained by solving it;a hand-eye calibration model was established to solve the coordinate system of the camera and the robot base.The transformation matrix is established;a dual-target fixed ranging model is established,and the calculation formula of the three-dimensional coordinates of the target point in the camera coordinate system of one station is constructed through the external parameters of the camera in the two stations.(2)Research on welding seam image processing algorithm.In order to extract the edge features of the crane girder weld clearly in the image,the main algorithms in the image processing are compared experimentally,and analyzed and discussed.The image preprocessing work is completed,which provides image information for subsequent weld feature identification,weld endpoint coordinate extraction and weld space coordinate calculation.(3)Calculation of weld space coordinates and error analysis.In order to obtain the spatial coordinates of the welding seam of the main beam of the crane,the Hough transform is used to obtain the edge pixel coordinates in the image of the edge of the welding seam,and the spatial coordinates of the welding seam endpoints are calculated.The error analysis of the spatial coordinates of the welding seam end points obtained by the vision system is completed.The relative error and distance error of the experimental results are within the technical requirements of the inner seam welding of the main beam of the crane,and the purpose of welding seam identification and positioning is achieved.The space coordinate value is imported into the welding robot program to complete the final welding experiment,and the coordinates of the identification and positioning of the inner seam of the crane main girder made in the experiment are further verified to be effective,and the welding effect is good.Through the research of this topic,a single-camera dual-station visual calibration model for the inner seam welding of the crane main girder is established,and the image processing algorithm for image feature extraction and coordinate calculation of the crane main girder weld image is studied,and the vision of the crane main girder welding seam is realized.Identification and positioning have great practical significance for improving the intelligent welding of the inner seam of the main beam of the crane.