基于机器视觉的钢制压力容器焊缝缺陷X射线检测方法  

X-ray Detection Method for Weld Defects of Steel Pressure Vessels Based on Machine Vision

作　　者：宋伟斌 赵建政 SONG Weibin;ZHAO Jianzheng(Yantai Special Equipment Inspection and Research Institute,Yantai 265508,China)

机构地区：[1]烟台市特种设备检验研究院,山东烟台265508

出　　处：《机械制造与自动化》2024年第5期242-246,共5页Machine Building & Automation

摘　　要：为降低由钢制压力容器焊缝缺陷引起的事故发生概率,提出基于机器视觉的钢制压力容器焊缝缺陷X射线检测方法。使用机器视觉系统以及X射线技术,获取钢制压力容器焊缝图像并提取焊缝区域特征,将焊缝缺陷整合成熔焊、虚焊、焊穿三大类别;基于焊缝区域特征的提取结果,提取钢制压力容器的焊缝缺陷面积、缺陷亮度以及波形等缺陷特征,构建缺陷分类器,实现钢制压力容器焊缝缺陷X射线检测。实验结果表明:钢制压力容器焊缝缺陷检测时,随着待检测焊缝图像的增加,F1值较高,实际缺陷检测结果的主观判断效果明显,细节检测效果好。An X-ray detection method of weld defects of steel pressure vessels based on machine vision is proposed to reduce the probability of accidents caused by weld defects of steel pressure vessels.Machine vision system and X-ray technology are applied to obtain the weld image of steel pressure vessel and extract the weld area characteristics,and the weld defects are integrated into three categories,fusion welding,false welding and penetration welding.Based on the extraction results of weld area features,the defect features such as weld defect area,defect brightness and waveform of steel pressure vessels are extracted,and the defect classifier is constructed to realize the X-ray detection of weld defects of steel pressure vessels.The experimental results show that when detecting the weld defects of steel pressure vessels,with the increase of the weld images to be detected,the F 1 value is higher,the subjective judgment effect of the actual defect detection results is obvious,and the detail detection effect is favourable.

关 键 词：机器视觉 X射线 钢制压力容器 焊缝缺陷 检测方法 

分 类 号：TG115[金属学及工艺—物理冶金]