The Coordinate Acquisition of the Sine Welding Seam based on Zernike Moments Edge Detection

The coordinate acquisition of curve seam in robotic laser welding is becoming very important due to its strict demands for high accuracy and high speed. The precision of the coordinates of curve seam is crucial to guarantee the welding quality. This paper presents an approach by using Ghosal's algorithm based on orthogonal Zernike moments for acquiring the coordinates of welding seam which is sine wave in shape. Firstly, two edges of the seam are detected by the edge detection algorithm based on Zernike moments. Then, by using median algorithm, a new sine curve connected with a string of single-pixel can be generated from the two detected seam edges. Secondly, the sub-pixel level coordinates of the sine seam are extracted. In the end, experiments have been implemented and results show that the approach is feasible.