Detecting and reconstructing curve welding seam using structured light stereovision

This paper mainly studies the approach of detecting and reconstructing curve butt welding seam based on a line structured light stereovision sensor. The approach consists of three steps: firstly, extracting the laser stripe center line from welding seam image, and determining the position of welding seam point; then, calculating its three-dimensional coordinates according to the calibration of vision sensor and hand eye matrix; and finally, reconstructing the complete curve welding seam in three-dimensional space by a group of welding seam points. Experimental results demonstrate the effectiveness of the proposed approach, which has great application value in robot automatic welding area.