Method for Identifying Arc Length of TIG Welding Based on Machine Vision and Image Processing

Abstract

In this paper, the TIG welding arc image is acquired by CCD camera, and the image processing technology is used to analyze and process the image to identify the length of welding arc. In this study, the pixel calibration method based on gradient operator and morphological operation is determined, an inflection detection algorithm that can accurately identify tungsten tip and an adaptive threshold algorithm that can obtain the best threshold to accurately segment the image is proposed. The experimental results show that this method can effectively identify the arc length, which is of great significance to the later study of visual-based arc length control method.