Conference paper
Weld pool visual sensing without external illumination
Visual sensing in arc welding has become more and more important, but still remains challenging because of the harsh environment with extremely strong illumination from the arc. This paper presents a low-cost camera-based sensor system, without using external Illumination, but nevertheless able to sense and model the weld pool.
Central is a carefully selected optical filtering as well as an active contour-based tracking of the weld pool boundary. The system is able to extract the 2D shape of the weld pool in real time. The reported experiments show the feasibility of this approach.
Language: | English |
---|---|
Publisher: | IEEE |
Year: | 2011 |
Pages: | 145-150 |
Proceedings: | 2011 IEEE International Conference on Automation Science and Engineering |
ISBN: | 1457717301 , 145771731X , 1457717328 , 9781457717307 , 9781457717314 and 9781457717321 |
ISSN: | 21618070 and 21618089 |
Types: | Conference paper |
DOI: | 10.1109/CASE.2011.6042506 |
ORCIDs: | 0000-0002-4466-1914 |
2D shape extraction Active contours Automation Computer Science Computer vision Control Engineering Faculty of Science Force Image Group Image edge detection Machine vision Real time systems Visualization Welding active contour-based tracking arc welding filtering theory image analysis image sensors low-cost camera-based sensor system object tracking optical filtering production engineering computing weld pool visual sensing