Unsupervised Recognition and Characterization of the Reflected Laser Lines for Robotic Gas Metal Arc Welding

Abstract

Unsupervised recognition of the reflected laser lines from the arc-light-modified background is prerequisite for the subsequent measurement and characterization of the weld pool shape, which is of great importance for the modeling and control of robotic arc welding. To facilitate the unsupervised recognition, the reflected laser lines need to be segmented as accurate as possible, which requires the segmented laser lines to be as continuous as possible to decrease the adverse effect of the noise blobs. In this paper, the intensity distribution caused by the arc light in the captured image is modeled. Based on the model, an efficient and robust approach is proposed, and it comprises six parts: reduction of the uneven image background by a difference operation, spline enhancement to remove the fuzziness, a gradient detection filter to eliminate the uneven background further, segmentation by an effective threshold selection method, removal of the noise blobs adaptively, and clustering based on the online computed slope of the laser line. After the laser line is clustered, a second-order polynomial is fitted to it. Finally, the weld pool is characterized by the parameters of the clustered laser line and its fitted polynomial. Experimental results verified that the proposed approach for unsupervised reflected laser line recognition is significantly superior to the state-of-the-art approach in terms of recognition accuracy.

Single-shot three-dimensional reconstruction based on structured light line pattern

“…In this research work, the quality of the captured images is much higher than that of [21][22] . Hence, we simplify the previously proposed segmentation approach and only select the gradient feature detection and the threshold selection to segment the bright lines for computation efficiency.…”

“…The segmentation approach proposed in [22] contains five parts: difference operation, spline fitting, gradient feature detection, threshold selection and filtering noise blobs because the quality of the captured laser line images is extremely low. In this research work, the quality of the captured images is much higher than that of [21][22] .…”

“…4 (a) shows the captured calibration grid image. Firstly, the calibration grid image is segmented by the slope difference distribution based threshold selection method [22] . The segmentation result is shown in Fig.…”

“…To calculate the distortions of the bright lines accurately, robust image processing algorithms are essential. We utilize the image processing methods proposed in the past research [21][22] to segment and cluster the bright lines from the captured image. After clustering, the skeleton of each clustered line is obtained and then smoothed by spline fitting.…”

Image Segmentation Approaches for Weld Pool Monitoring during Robotic Arc Welding

“…In the past decades, a lot of research has been conducted to measure the shape of the arc welding weld pool by structured light methods [9,[15][16][17][18][19][20][21][22][23]. In Reference [15], the authors pioneered to measure the deformation of the weld pool by inventing a novel sensing system.…”