摘要
针对射线检测人工评片方式主观性强、效率低、稳定性差,以及底片资料不易保存、信息管理工作较落后等缺点,本文通过利用人工智能的深度学习方法自动提取图像深层信息的特质,基于大数据开发射线检测底片识别系统,开启无损检测信息化大数据新模式,实现筛选无效底片与重复底片、焊缝缺陷位置识别等智能评片功能,完成底片管理信息报告的自动生成保管工作,对提高特种设备行业无损检测的安全性和可靠性具有重要意义。
In view of the shortcomings of the manual evaluation method for ray detection,such as strong subjectivity,low efficiency,poor stability,as well as the difficulty in storing negative data,and backward information management,the deep learning method of artificial intelligence was used to automatically extract the characteristics of deep information in images,a radiographic testing negative recognition system based on large data was developed,a new mode of non-destructive testing information big data was opened,and the screening of invalid and duplicate negative iden Intelligent film evaluation functions was realized,Intelligent film evaluation functions such as screening invalid and duplicate negative films,and identifying the location of weld defects,as well as completing the automatic generation and storage of negative management information reports,which was of great significance for improving the safety and reliability of non-destructive testing in the special equipment industry.
作者
陈俊仰
邓聪
李绪丰
郑俊辉
李明飞
CHEN Jun-yang;DENG Cong;LI Xu-feng;ZHENG Jun-hui;LI Ming-fei(Guangdong Special Equipment Inspection and Research Institute,Guangdong Foshan 528251,China)
出处
《广州化工》
CAS
2023年第2期19-21,35,共4页
GuangZhou Chemical Industry
基金
国家市场监督管理总局科技计划项目(2021MK087)。
关键词
射线检测
人工智能
底片识别
信息化
ray detection
artificial intelligent
negative identification
information
