焊缝X射线图像的数学形态学分析

Mathematical Morphology Analysis of Weld X-ray Images

焊接过程中产生的缺陷影响焊缝质量,X射线无损探伤是常用的检测技术手段。针对焊缝X射线图像的缺陷提取问题,提出使用数学形态学分析的技术方案。先用低帽变换处理原始图像,降低背景灰度的同时凸显缺陷区域,然后设置固定阈值分离缺陷区域,最后使用开运算去除干扰,实现缺陷的自动提取。实验结果表明,提出的技术方案合理可行,能够准确分离焊缝X射线图像上的缺陷区域。

Defects formed in welding process affect the quality of weld seam,and X-ray nondestructive testing is a commonly used testing technology.Aiming at the problem of defect extraction from X-ray images of weld seams,a technical scheme using mathematical morphology was proposed.Firstly,the original image was processed by bottom hat transformation,which could reduce the background gray level and highlight the defect area.Then a fixed threshold was set to separate the defect area.Finally,the open operation was used to remove the interference and realize the automatic defect extraction.The experimental results show that the proposed technical scheme is reasonable and feasible,which can accurately separate the defect area in the X-ray image of weld seam.