Charaterization Of 3D Outline Groove And Weld Surface Based On Laser Vision

With the application of robot technology and digital manufacturing in the equipment manufacturing industry, welding automation also has a rapid development. The welding process is more intelligent and reliance on welders is more low. At present, the SAW welding process relies on the human experience of neutralization welder of path teaching, which is a waste of time and energy.The heat input of the large line energy leads to the distortion of the workpiece to be welded, and the path of the welding deviation before the welding is deviated.As a result, the welding process of submerged arc welding often appears to be a serious defect, such as welding deviation, incomplete penetration and so on. In view of the above problems, a laser vision measurement system was set up. A method was provided to get the 3D surface outline base on the laser visual measurement system. The main content is shown as following:(1) A laser visual measurement system was set up. The laser stripe scanned groove by relative motion between the workpiece and laser vision measurement system, then got a series of images can reflect the changes of groove and weld profile. The relationship between the pixel coordinate system and the world coordinate system is determined by the calibration of the laser vision measurement system, so that the correspondence between the pixel size and the physical size of the workpiece can be determined. The laser vision system can complete the groove and weld contour information acquisition(2) Weld groove and visual features were analysed and treated. Laser line can well reflect the groove and the weld contour changes through the known correlation of fringe analysis and groove and welding seam contour feature.At the same time,the laser line follow the trend of changes of stripe groove and weld profile. Therefore the laser line can be used as characterization of groove and weld contour information source.(3) Realize the coordinates of discrete points on laser line. Through the enhancement of anisotropic diffusion filtering, the maximum between class variance method and region growing method is used to realize laser line segmentation. Finally, the laser line of single pixel width is obtained. The method to obtain the coordinates of discrete points on the fringes is established,and the coordinates of discrete points on the fringes are realized. Laser line and scanning of the groove and weld profile section is one-to-one, so we can obtain any cross section point coordinates from continuous scanning by the laser line.(4) The digital reconstruction and groove weld surface profile was realized.In order to overcome the problem of measuring blind zone caused by the moving speed matching between the camera and the welding trolley, a searching fitting interpolation algorithm is proposed to obtain the coordinates of the points of the stripes. The neighbor points was connected into the grid method to reconstruct the 3D digital outline. Experimental verification shows that any section obtained from digital contour on the groove width and depth with vernier caliper to measure the value of more than 98%. The 3D digital outline of weld groove can replace the manual measurement results.(5)The digital representation parameters and obtaining methods of 3D profile are established. According to the requirement of welding filling strategy and path planning, the geometric characteristic parameters were established,such as groove width Wi, groove depth profile Hi, section Si, the welding trajectory curve of Z. Based on the above parameters, the filling quantity is determined. It can provide basic data for welding process and quality assurance of the intelligent based on laser vision measurement structure and weld groove3 D digital outline.