Study on welding seam tracking for robotic arc welding under strong surface reflection

Purpose This study aims to improve the environmental adaptability of teach-and-playback welding robots and the ability of laser vision sensing-based robotic seam tracking technology to cope with strong workpiece surface reflections and short lead distance between sensor and welding torch. Welding seam tracking algorithms have become a core research objective for promoting the stable application of robotic welding technology in complex welding scenarios. Design/methodology/approach This paper proposed an oblique-incidence laser vision sensing robotic welding seam tracking algorithm suitable for strong surface reflection conditions. First, to recover laser stripe feature from noisy images, a spatial coordinate attention module and dual-scale fusion module were developed to construct an image segmentation network. By fully leveraging contextual and multiscale feature information, this network enabled more precise restoration of laser stripes under large-area noise occlusion. Then, to further improve robustness, a self-iterative random sample consensus algorithm was developed to mitigate the effects of erroneous laser stripe inputs, facilitating accurate localization of welding seam position. Finally, a weld seam tracker was constructed. Findings The effectiveness of the proposed algorithm was validated on four workpieces with diverse surface conditions and laser stripe image data sets contaminated by large-area noise. Experimental results demonstrate that the proposed tracker reduced the error in weld seam position pixel extraction accuracy by 81%, achieved a weld seam position pixel extraction accuracy of 0.52 pixels, and operated at an average processing speed of 37 FPS, under 22 mm lead distance. Practical implications The proposed approach could be applied to the robotic welding fabrication of products such as container girders, marine vessels, pressure vessels and vehicles. Originality/value This study addresses the persistent challenge in industrial practice whereby vision-sensing robotic welding technologies are prone to tracking failure under conditions of strong workpiece surface reflections and short lead distance between the sensor and the welding torch, thereby filling a critical research gap and offering a technical approach.

Authors

Institutions

Publication Details

Journal
Industrial Robot the international journal of robotics research and application
Published
2026-11-16
DOI
https://doi.org/10.1108/ir-05-2026-0245
Primary Topic
Industrial Vision Systems and Defect Detection
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Study on welding seam tracking for robotic arc welding under strong surface reflection

JS Gao, Kai Wang, Yi Zheng, Yuanyuan Qu et al.
1 links
Industrial Robot the international journal of robotics research and application
Industrial Vision Systems and Defect Detection
article

Study on welding seam tracking for robotic arc welding under strong surface reflection

JS Gao, Kai Wang, Yi Zheng, Yuanyuan Qu, Hao Wang
article en

Abstract

Purpose This study aims to improve the environmental adaptability of teach-and-playback welding robots and the ability of laser vision sensing-based robotic seam tracking technology to cope with strong workpiece surface reflections and short lead distance between sensor and welding torch. Welding seam tracking algorithms have become a core research objective for promoting the stable application of robotic welding technology in complex welding scenarios. Design/methodology/approach This paper proposed an oblique-incidence laser vision sensing robotic welding seam tracking algorithm suitable for strong surface reflection conditions. First, to recover laser stripe feature from noisy images, a spatial coordinate attention module and dual-scale fusion module were developed to construct an image segmentation network. By fully leveraging contextual and multiscale feature information, this network enabled more precise restoration of laser stripes under large-area noise occlusion. Then, to further improve robustness, a self-iterative random sample consensus algorithm was developed to mitigate the effects of erroneous laser stripe inputs, facilitating accurate localization of welding seam position. Finally, a weld seam tracker was constructed. Findings The effectiveness of the proposed algorithm was validated on four workpieces with diverse surface conditions and laser stripe image data sets contaminated by large-area noise. Experimental results demonstrate that the proposed tracker reduced the error in weld seam position pixel extraction accuracy by 81%, achieved a weld seam position pixel extraction accuracy of 0.52 pixels, and operated at an average processing speed of 37 FPS, under 22 mm lead distance. Practical implications The proposed approach could be applied to the robotic welding fabrication of products such as container girders, marine vessels, pressure vessels and vehicles. Originality/value This study addresses the persistent challenge in industrial practice whereby vision-sensing robotic welding technologies are prone to tracking failure under conditions of strong workpiece surface reflections and short lead distance between the sensor and the welding torch, thereby filling a critical research gap and offering a technical approach.

Industrial Robot the international journal of robotics research and application
Hunan University of Science and Technology (CN), Anhui University of Technology (CN)
Industrial Vision Systems and Defect Detection
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon