Weld Seam Tracking Based on Laser Imaging Binary Image Preprocessing

摘要

In the current welding seam tracking research, the laser tracking system of welding seam in the v-shaped welding process, the camera’s anti-interference ability is low in the process of image acquisition, image noise is too much, and so on. A welding seam tracking method of laser imaging binary image preprocessing is proposed. The denoising point method of open operation and OSTU algorithm were used to obtain image binarization threshold and generate binarization image. MATLAB programming was used to obtain the binary image of the laser structural light obtained by the camera, and the image was further de-noised by corrosion and expansion operations to obtain the accurate data of the weld. The laser imaging features in the V-shaped weld image were enhanced to further improve the identification accuracy of using the laser vision triangulation method, as well as the tracking accuracy of the weld during the process. The error could be controlled within 0.5mm.