摘要
采用微焦点X射线检测系统获取T形接头焊件左、右旋转的检测图像,提取感兴趣区域,分析图像灰度分布特征,采用小波变换对感兴趣区域进行图像去噪处理。采用形态学背景模拟,结合阈值分割等算法实现了图像中缺陷的保真分割。建立T形接头焊件缺陷深度和缺陷偏移量的数学模型,提出了焊件缺陷投影距离的自动提取算法及左、右旋转图像中缺陷的自动对应准则,自动提取T形接头焊件内部缺陷空间位置数据。试验结果显示,缺陷大多分布在腹板中心线的左侧,且集中分布在焊件腹板和翼板结合面附近的位置,缺陷沿焊缝纵向呈随机分布,缺陷半径范围为[0,0.7]mm。
The micro focus X-ray detection system is used to obtain the left and right rotation detection image of T-joint weldment,extract the region of interest,analyze the gray distribution characteristics of the image,and use wavelet transform to denoise the image of the region of interest.The fidelity segmentation of defects in images is realized by using morphological background simulation and combining threshold segmentation algorithms.The mathematical model of defect depth and defect offset of T-joint weldment is established,and the automatic extraction algorithm of welding defect projection distance and the automatic corresponding criterion of defects in left and right rotating images are proposed,so as to automatically extract the spatial position data of internal defects of T-joint weldment.The experimental results show that most of the defects are distributed on the left side of the centerline of the web,and are concentrated near the junction surface of the web and wing of the welded part.The defects are randomly distributed along the longitudinal direction of the weld seam,with a defect radius range of[0,0.7]mm.
作者
孙远
石端虎
SUN Yuan;SHI Duanhu(School of Mechanical Engineering,Yancheng Institute of Technology,Yancheng Jiangsu 224051,China;School of Mechanical and Electrical Engineering,Xuzhou University of Technology,Xuzhou Jiangsu 221018,China)
出处
《盐城工学院学报（自然科学版）》
CAS
2023年第2期25-31,共7页
Journal of Yancheng Institute of Technology:Natural Science Edition
基金
江苏省自然科学基金面上项目(BK20141143)
江苏省高校自然科学研究基金重大项目(16KJA430003)
江苏省青蓝工程优秀教学团队资助项目(2021)。
关键词
T形接头焊件
图像去噪
缺陷分割
空间位置数据
自动提取
T-shaped weldment
image denoising
defect segmentation
spatial position data
automatic extraction
