Home > Published Issues > 2020 > Volume 9, No. 2, February 2020 >

A Study on Development of Seam Tracking Algorithm in Robotic GMA Welding Process

Gang Zhang, Tae-Jong Yun, Won-Bin Oh, Bo-Ram Lee, and Ill-Soo Kim
Department of Mechanical Engineering, Mokpo National University, Mokpo, South Korea

Abstract—In weld seam tracking system, image processing plays an important role in obtaining accurate weld information. Due to the large number of noise signals in the weld image, the surface condition of the weldment and the illumination environment also have a great influence on the image processing results.

This paper develops an image pre-processing algorithm that based on thermal high-speed camera, which mainly includes image noise removal algorithm and contrast enhancement algorithm. In the noise removal algorithms, three kinds of noise filtering (minimum filtering, median filtering and maximum filtering) were employed. In addition, four morphological operators (erosion, dilation, opening and closing operation) were utilized in the image contrast enhancement processing. The proposed algorithms are validated and compared to obtain an optimal algorithm for each image processing step. The simulated results show that the median filtering algorithm and the closing operation are the preferred methods because these algorithms provide lower RMSE (Root Mean Square Error) and higher PSNR (Peak Signal-to-Noise Ratio). Therefore, median filtering was applied to reduce the noise of the seam image, and closing operation was used for image contrast enhancement. Finally, the threshold is obtained to binarize the image to obtain a better enhancement effect based on the Otsu’s method. 
 
Index Terms— GMA welding, Noise filter, Contrast enhancement algorithm, Image pre-processing

Cite: Gang Zhang, Tae-Jong Yun, Won-Bin Oh, Bo-Ram Lee, and Ill-Soo Kim, "A Study on Development of Seam Tracking Algorithm in Robotic GMA Welding Process" International Journal of Mechanical Engineering and Robotics Research, Vol. 9, No. 2, pp. 310-313, February 2020. DOI: 10.18178/ijmerr.9.2.310-313

Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.