基于视频流的连锁信号智能识别系统设计与实现  

Intelligent Recognition System for Chain Signals Based on Video Streaming

作　　者：任石 Ren Shi

机构地区：[1]中国铁路沈阳局集团有限公司运输部,辽宁沈阳110000

出　　处：《工业控制计算机》2024年第11期4-5,8,共3页Industrial Control Computer

基　　金：中国铁路沈阳局科技项目(2022Y03:无线调车灯显设备与本务机车“LKJ”配合使用的研究)资金支持。

摘　　要：针对铁路运维中计算机连锁系统信号采集不便、摄像头拍摄方式易受干扰、识别误差大等问题,设计了一种基于视频流的连锁信号智能识别系统。通过VGA分配器分出一路视频,经视频编码器形成标准视频流,然后交给智能主机进行连锁信号的区域类型匹配及智能分析识别。结合实际业务逻辑,将识别结果自动组合生成客户需要的铁路调度及维护的专业话术用语,供现场工作人员审核、修改和发布。实际运行表明,基于视频流的连锁信号智能识别系统识别速度快、准确度高、抗干扰能力强,具有重要的理论意义和实用价值。A video stream based intelligent recognition system for chain signals is designed to address the issues of inconvenient signal acquisition and large error camera acquisition in railway operation and maintenance.A channel of video information is separated through a VGA splitter,and a standard video stream is formed by a video encoder.It is then handed over to an intelligent host for region type matching and intelligent analysis and recognition of chain signals.Based on actual business logic,the recognition results will be automatically combined to generate professional language statements for railway scheduling and maintenance,for on-site staff to review,modify,and publish.The intelligent recognition system for chain signals based on video streams has fast recognition speed,high accuracy,and strong anti-interference ability.

关 键 词：视频流 连锁信号 类型匹配 智能识别 

分 类 号：U284.3[交通运输工程—交通信息工程及控制] TP391.41[交通运输工程—道路与铁道工程]