A Submerged Arc Weld Detection Technology Based On Light Coding Technology

Vision sensors and image processing technology are widely used in the ifeld of weldingprocess control and seam tracking. A set of weld inspection methods and automatic alignmentsystem as the core of depth image processing was designed上y the research of light codingtechnology vision sensors and depth image processing method in submerged arc welddetection.The performance of the light coding-based depth sensor has analyzed by experiment andstudied the application of depth sensor technology in automatic welding system. VS2010isused to develop the depth image detection and processing techniques, which include the depthimage acquisition, depth image processing, image display and control signal output. Thesystem has the main function of video display and saving. ROI region display, ifltering methodselection, edge detection algorithm selection, algorithm parameter selection as well as thedetection of the weld center and weld center iftting.Discussed a submerged arc weld detection algorithm based on the depth image, it cancorrectly identify the submerged arc weld center by finding the maximum depth value for eachrow ofpixels directly without sophisticated image pre-processing. After a median filtering andthe center point fitting of the weld image, accurate and stable weld curve center has beengotten.Experiments show that weld inspection system required low light environment and imagepre-processing, fewer processing steps, simple algorithm. V-weld centerline can be detectedsteadily, rapidly and accurately, provide a practical and effective feedback signal to achievethe automatic control of the torch position.