EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Deep Learning-Based Framework for Offline Robotic Weld Path Generation Using a Single Top-View RGB-D Image

Dahyeon Lee iD, Byungjin Ko, Taejoon Park iD, Jong-Wan Yoon iD, Homin Park iD

Sensors · 05.08.2026 · 26 (15) · 4973

DOI: 10.3390/s26154973 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 05.08.2026. Verfasst von Dahyeon Lee und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Applied Artificial Intelligence, Hanyang University, Ansan 15588, Republic of Korea. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Government of the Republic of Korea, Korea Institute of Industrial Technology, Korean government.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Pipe welding automation requires accurate weld seam extraction and reliable robotic weld path generation under complex geometric conditions. Existing vision-based approaches often rely on expensive laser sensing systems, multi-view sensing, or continuous seam tracking, resulting in increased hardware cost and system complexity. To address these limitations, this study proposes a deep learning-based offline robotic welding framework that generates a three-dimensional welding path from a single top-view Red–Green–Blue and Depth (RGB-D) image acquired prior to welding. The proposed framework integrates weld seam detection, semantic segmentation, morphology-based post-processing, RGB-D image alignment, coordinate transformation, and polynomial-based trajectory refinement into a unified pipeline for robotic weld path generation. A custom pipe welding dataset consisting of 1476 annotated images collected from representative industrial pipe materials with varying diameters was constructed to evaluate the proposed framework. The experimental results demonstrate that the proposed Region of Interest (ROI)-guided weld seam extraction pipeline improves the U-Net segmentation performance from 0.735 to 0.791 mean Intersection over Union (mIoU), while the detection model achieves a Recall of 0.988 and an mean Average Precision at an Intersection over Union threshold of 0.5 (mAP50) of 0.995. Furthermore, polynomial-based trajectory refinement reduces the three-dimensional positional root mean square error (RMSE) to 0.333 mm, enabling continuous robotic welding over the entire visible weld seam without additional path modification. These results demonstrate that the proposed framework provides a practical and cost-effective solution for offline robotic weld seam extraction and weld path generation, while establishing a promising foundation for future extension toward online robotic welding through real-time weld seam tracking and adaptive trajectory correction.

Schlagwörter

Stichwörter der Autor:innen

RGB-D sensorcomputer visionpipe weldingrobot weldingrobotic automationsegmentationweld seam extraction

Autor:innen und Institutionen

  • Department of Applied Artificial Intelligence, Hanyang University, Ansan 15588, Republic of Korea.
  • Division of Smart Convergence Engineering, Hanyang University-ERICA, Ansan 15588, Republic of Korea.
  • Department of Robotics Engineering, Hanyang University-ERICA, Ansan 15588, Republic of Korea.
  • Department of Intelligent Robotics, Hanyang University-ERICA, Ansan 15588, Republic of Korea.
  • School of Science and Technology, Singapore University of Social Sciences, Singapore 599494, Singapore.

ORCID-Profile: Dahyeon Lee, Taejoon Park, Jong-Wan Yoon, Homin Park

Förderung

  • Government of the Republic of Korea
  • Korea Institute of Industrial Technology
  • Korean government

Zitieren

APA

Lee, D., Ko, B., Park, T., Yoon, J. W., & Park, H. (2026). A Deep Learning-Based Framework for Offline Robotic Weld Path Generation Using a Single Top-View RGB-D Image. Sensors, 26(15), 4973. https://doi.org/10.3390/s26154973

Vancouver

Lee D, Ko B, Park T, Yoon JW, Park H. A Deep Learning-Based Framework for Offline Robotic Weld Path Generation Using a Single Top-View RGB-D Image. Sensors. 2026;26(15):4973. doi:10.3390/s26154973

BibTeX
@article{Lee2026,
  author = {Lee, Dahyeon and Ko, Byungjin and Park, Taejoon and Yoon, Jong-Wan and Park, Homin},
  title = {A Deep Learning-Based Framework for Offline Robotic Weld Path Generation Using a Single Top-View RGB-D Image},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {15},
  pages = {4973},
  doi = {10.3390/s26154973}
}
Kostenlos starten →