隧道行车图像特征信息提取及综合应用方法研究

A Method Research for Feature Information Extraction and Application of Tunnel Vehicle Image

本文提出一种通过视频图像特征信息提取的方法,检测隧道中是否有车通过、行车速度、来车数量等信息,智能控制系统用其实施"车近灯亮、车过灯灭、依据车速自动调整隧道灯的照明亮度"。另外,将隧道行车信息提取算法中产生的数据合理储存,通过数据挖掘,1)依据车流量及车速判断隧道的交通状况,异常情况下主动报警,并向交管部门发送特征图像,供交通疏导或事故救援的指挥作参考;2)检索某人驾车何时通过该监测点;3)获取月度、季度及年度的通行量报表等。

This paper presents a method of processing video image by feature information extraction in tunnel intelligent control system.By detecting whether vehicles pass,their driving speed and vehicle number information in tunnels,the brightness scheme of the tunnel lights of ＂the car near the lights on,the car far the lights off＂can be implemented and adjusted automatically.In addition,the data generated by information extraction and data mining can be applied as follows.Firstly,the system can forecast tunnel traffic conditions judging by traffic flow and speed,alarm and send the information to the traffic control department when necessary,which is helpful to the tunnel traffic grooming or accident rescue.Secondly,it has the function to retrieve the time when a vehicle passes the monitoring point.Thirdly,the system can also provide monthly,quarterly or annual traffic volume of the tunnel etc.