A Seam Tracking Algorithm Utilizing Spatio-Temporal Information of Laser Vision in Thin Plate Butt Welding for Shipbuilding

Thin plates constitute more than 80% of the structural components in shipbuilding, and butt welding is an essential technological process ship manufacturing. Different from general welding, thin plate for shipbuilding characterized by long length (≥ 15 m), narrow seam width (≤ 1 mm), high precision demands 0.5 mm). The clamping deformation, arc sparks, noise pollution during seriously affect tracking automated welding. To address this problem, paper presents a algorithm combining image processing techniques spatiotemporal data filtering to acquire steady smooth torch trajectory with less position variation. In stage, density mask method proposed preserve effective pixels laser images resist noises improve accuracy point estimation. Experimental results show that frequency can reach 30 Hz, average absolute error 0.25 mm, maximum 0.4 mm. runs smoothly thus leading higher quality.