铁路通信信号工程技术2025，Vol.22Issue(9)：115-121,7.DOI:10.3969/j.issn.1673-4440.2025.09.019
图像识别技术在信号机械室智能巡检系统中的研究
Research on Image Recognition Technology in Intelligent Inspection Systems for Signal Equipment Rooms
摘要
Abstract
Image recognition technology enables the automatic identification of equipment operating conditions,facilitating intelligent inspections in signaling equipment rooms and,thus,enhancing inspection efficiency.This paper first provides an introduction to intelligent inspection systems and analyzes the application of image recognition technology within the mechanical room.It then introduces image recognition algorithms such as OpenCV,YOLO,and CRNN+CTC,and analyzes the advantages,disadvantages,and application scenarios of each algorithm.Finally,the paper proposes solutions to address challenges encountered during the practical application of image recognition technology,such as image displacement,environmental color differences,computational power consumption,and system setup,in order to improve image recognition accuracy.Pilot trials conducted at several railway bureaus,including those in Shanghai,Wuhan,Zhengzhou,and Xi'an,have significantly reduced the workload of daily field inspections,yielding positive results.关键词
图像识别技术/识别场景/YOLO/OpenCV/CRNN+CTC/智能巡检Key words
image recognition technology/recognition scenarios/YOLO/OpenCV/CRNN+CTC/intelligent inspection分类
交通工程引用本文复制引用
陈壁东..图像识别技术在信号机械室智能巡检系统中的研究[J].铁路通信信号工程技术,2025,22(9):115-121,7.基金项目
广州地铁集团有限公司重大专项(HT210228) （HT210228）
