Vision-based algorithm for online TIG welding deviation detection of stainless steel welded pipes.

Area of Science:

• Materials Science and Engineering
• Manufacturing Processes
• Computer Vision

Background:

• Tungsten Inert Gas (TIG) welding is crucial for stainless steel pipe production.
• Accurate monitoring of the welding process is essential for quality control.

Purpose of the Study:

• To develop an automated method for precise weld seam identification in TIG welding.
• To enable accurate calculation of welding deviations using image analysis.

Main Methods:

• Utilizing image moments to calculate the tungsten needle tip position from molten pool images.
• Extracting weld regions and interpolating contours using cubic B-spline curves.
• Applying Savitzky-Golay (S-G) filtering for noise reduction and feature point extraction.

Main Results:

• Successfully extracted weld feature points for accurate weld seam identification.
• Demonstrated the capability to precisely calculate welding deviations during the TIG welding process.
• The developed method shows high accuracy in identifying weld seams in stainless steel pipes.

Conclusions:

• The proposed image processing technique effectively identifies weld seams in TIG welding of stainless steel pipes.
• This method provides a reliable approach for real-time monitoring and control of welding deviations.
• Automated weld seam identification enhances the efficiency and quality of stainless steel pipe manufacturing.