X射线数字图像缺陷自动识别应用  

Automatic recognition application of X-ray digital image

作　　者：张甲鹏 李飞[1] 王占富[1] ZHANG Jiapeng;LI Fei;WANG Zhanfu(Harbin Boiler Company Limited,Harbin 150046,China)

机构地区：[1]哈尔滨锅炉厂有限责任公司,黑龙江哈尔滨150046

出　　处：《锅炉制造》2024年第4期38-39,共2页Boiler Manufacturing

摘　　要：针对目前无损检测结果评定主要采用人工方式,采用缺陷智能识别技术对X射线数字图像缺陷进行自动识别,通过深度学习功能模仿人脑并建立一个类似的学习策略,使缺陷智能识别技术辅助检测人员对检测图像进行评估,减少缺陷的漏检率,提高产品质量的一致性。At present,the evaluation of NDT results mainly adopts manual method.Intelligent defect recognition technology is used to identify defects in X-ray digital images.Using deep learning to mimic the human brain and build a similar learning strategy.The intelligent defect recognition technology is used to assist the inspector to evaluate the detection image.Reduce the rate of defect omission.Improve consistency of product quality.

关 键 词：X射线数字图像 深度学习 自动识别 

分 类 号：TH878[机械工程—仪器科学与技术]