计算机与数字工程2024，Vol.52Issue(4)：1216-1220,5.DOI:10.3969/j.issn.1672-9722.2024.04.046
融合语义理解的航站楼显示设备故障检测方法
A Fault Detection Method of Terminal Display Equipment Fusing Semantic Understanding
摘要
Abstract
A fault detection method that integrates semantic understanding is proposed to address the fault images of display equipment in terminal buildings.Firstly,a rolling text stitching technique is designed to extract interface text information.Then,based on the airport business background,semantic rules are integrated to train a fault classification model,which realizes intelli-gent detection of abnormal display interfaces and display information ambiguities in display devices.Finally,the neural network model compression technology is used to lightweight the model and deploy it on SOM-RK3399 embedded devices.The experiment shows that the classification accuracy of the detection method integrating semantic understanding module reaches 88.74%.This method can effectively solve the shortcomings of traditional fault detection techniques,improve fault detection efficiency,and re-duce manual inspection situations.关键词
语义理解/滚动文字拼接/故障检测/模型轻量化Key words
semantic understanding/rolling text stitching/failure detection/model lightweight分类
信息技术与安全科学引用本文复制引用
张丹,潘芙兮,李光耀..融合语义理解的航站楼显示设备故障检测方法[J].计算机与数字工程,2024,52(4):1216-1220,5.基金项目
中国民航大学研究生科研创新项目(编号:2020YJS031)资助. （编号:2020YJS031）
