An Edge Detection Method for Welding Pool Based on An Improved Canny Algorithm
Lin Zou,Yunlong Zheng,Jie Lu
DOI: https://doi.org/10.1088/1742-6596/2785/1/012013
2024-07-04
Journal of Physics Conference Series
Abstract:In this study, an improved Canny algorithm is proposed for detecting welding pool edges in complex environments, considering the issues of inaccurate extraction, discontinuity, and susceptibility to noise in traditional edge detection algorithms. Firstly, the original Gaussian filter is replaced with a hybrid filter that combines dark channel prior defogging and bilateral filtering used for image denoising. Then, a four-direction Sobel operator is utilized to calculate gradient amplitude and direction. Furthermore, the double threshold required for extracting and connecting pool edges is obtained by using an enhanced quadratic Otsu algorithm. Experimental results demonstrate that, compared with the traditional algorithm, the improved Canny algorithm achieves an approximately 51.6% increase in PSNR and about 0.04 improvement in SSIM, enabling more accurate and comprehensive extraction of welding pool edge information.