Reproduce “Automated Weld Seam Recognition and 3D Mapping for Robotic Post Processing Using Photogrammetry and Semantic Segmentation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Augustin Raju, Abilash Madavath, Chandra Yuvesh Aubeeluck, Nicolas Pyschny, Felix Hackelöer, Florian Zwanzig
Researchers at TH Köln developed an automated vision-based pipeline to map weld seams in 3D for robotic post-processing, utilizing photogrammetry with standard cameras and semantic segmentation. The system provides approximate weld seam localization on large workpieces with a Root Mean Square Error of 4.2 mm, demonstrating its potential for guiding high-precision robotic operations.
Key insights
- Readily available hardware like smartphone cameras, combined with photogrammetry, can generate sufficiently accurate approximate 3D workpiece models.
- Semantic segmentation can effectively identify weld seams in 2D images across varying conditions, forming a robust input for 3D projection.
- Integrating these vision-based techniques creates a viable and efficient preliminary 3D localization system for guiding subsequent robotic tasks on large components.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.