地铁列车障碍物视频识别系统设计

Design of Video Recognition System for Subway Train Obstacle

基于视频图像识别技术的列车障碍物识别系统,可以实现无人驾驶列车前进方向轨道内障碍物的自动检测。该系统通过采用2台独立的高清摄像机采集图像数据,经软件的智能计算分析,可识别列车前方是否存在障碍物,并实现障碍物与车辆之间的距离测量。该技术在业内属首次开发,可实现列车前方障碍物的检测,有效替代司机进行线路瞭望。目前本系统已通过功能仿真验证测试,拟在实际运行线路的列车上进一步进行功能验证。

The train obstacle recognition system based on video recognition technology can realize automatic detection of obstacle on the track in front of the driverless train.Through analyzing the image data collected by two indepen-dent high definition cameras with intelligent computing soft-ware,obstacles in front of the train are detected,the distance between the obstacle and vehicle is measured. This technolo-gy is first developed in rail transit industry to detect obstacles in front of the train and effectively replace the driver′ s line look. At present,the system functional simulation has passed-verification test,further verification of the system function is proposed on the trains running on actual line.