التفاصيل البيبلوغرافية
| العنوان: |
Research on robust laser vision feature extraction method for fillet welds with different reflective materials under uncertain interference. |
| المؤلفون: |
Wu, Chaoqun (AUTHOR), Hu, Jiahao (AUTHOR), Lei, Ting1 (AUTHOR) leiting0621@whut.edu.cn, Yang, Peiwen (AUTHOR), Gu, Shiyang (AUTHOR) |
| المصدر: |
Optics & Laser Technology. Feb2023:Part A, Vol. 158, pN.PAG-N.PAG. 1p. |
| مصطلحات موضوعية: |
*REFLECTIVE materials, *CORNER fillets, *FEATURE extraction, *LASER welding, *CARBON steel, *WELDING equipment |
| مستخلص: |
[Display omitted] • Weld feature extraction of fillet weld in different reflective materials is taken into consideration. • The reflective interference can be overcome by the method of dual-slope judgment and double least squares fitting. • A prior morphological judgment method is proposed to filter out uncertain noise when welding. Welding seam tracking is an important means to improve welding quality and realize intelligent welding, among which, laser vision feature extraction for weld seam is one of the important research fields of welding seam tracking. However, current laser vision weld feature extraction methods often target single-material weldments. For weldments of different materials (different reflections), the generality and robustness of the existing methods are insufficient. To address this problem, the fillet welds of different reflective materials such as carbon steel (weakly reflective material), stainless steel (medium reflective material) and aluminum alloy (strongly reflective material) are used as research objects, and the feature extraction method of fillet welds of different reflective materials under interference is studied. According to the morphological characteristics of noise in laser stripe images, the noises are divided into four types. A prior morphological judgment method (PMJM) is proposed to filter out these four kinds of noise. The method is mainly divided into four steps, each step removes one kind of noise, and finally completes the removal of all noises. Since PMJM is greatly affected by the extraction accuracy of the initial frame, a fillet weld feature extraction method based on dual-slope judgment and double least squares fitting is proposed, which can accurately extract the characteristics of the fillet weld in the first frame when there has no arc noise. The experimental results show that the error of the proposed method is within ± 0.13 mm in the X-axis and Y-axis directions under three different reflective materials, and the average computation time of the three materials is 88 ms. The results show that the method has the advantages of strong robustness and high precision, and can meet the requirements of real-time welding. [ABSTRACT FROM AUTHOR] |
| قاعدة البيانات: |
Academic Search Index |
