Color Image Segmentation of Plume for Monitoring Disk Laser Welding Process

Abstract:

During deep penetration laser welding, changes in the metal vapor plume contain information about the stability of welding process. A high-speed camera was used to online monitor the welding process in order to detect the laser welding defects. A color segmentation clustering algorithm based on HSI color space was proposed for processing the recorded welding sequences. The effectiveness of algorithms based on different model is discussed, welding experimental results showed that the proposed algorithm could achieve better image segmentation, and it highlighted the edge of the metal vapor details in the image.

• Periodical: Applied Mechanics and Materials (Volumes 201-202)
• Pages: 166-169
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.201-202.166
• Online since: October 2012
• Authors: Teng Wang, Xiang Dong Gao
• Keywords: Color Segmentation Clustering Algorithm, Disk Laser Welding, HSI Color Space

References

[1] G.H. Li, Y. Cai and Y.X. Wu: Opt. Lasers Eng., Vol. 47 (2009), pp.990-994.

[2] D.Y. You, X.D. Gao, S. Katayama: Advanced Materials Research, Vol.314-316 (2011), pp.937-940.

[3] D.C. Tseng and C.H. Chang: Proc. 11th Int. Conference on Pattern Recognition (Netherlands, Hague, August 30-September 3, 1992), pp.228-231.

[4] A. Koschan and M. Abidi: Digital Color Image Processing (Wiley Publications, United States of America 2008).