基于监控视频分析的高速公路能见度检测与预警系统研究

发布时间：2018-04-18 19:33

本文选题：高速公路 + 视频监控　；参考：《长安大学》2016年硕士论文

【摘要】：近些年来,由于自然环境的恶化,由雾霾、沙尘、强光等引起的低能见度天气对我国高速公路安全行车的影响越来越为严重,高速公路沿线的能见度天气全面检测与预警处置需求也变得愈来愈迫切。相比于常规能见度测量仪器造价昂贵、采样空间小、安装复杂、维护困难等缺点,密布于高速公路沿线的监控摄像机电设备所记录的交通监控视频包含了丰富的能见度信息,为高速公路能见度探测提供了新的思路。有鉴于此,本文利用高速公路沿线监控摄像头拍摄的监控视频,设计了一种基于监控视频的能见度检测方法,并围绕该方法开发了一款分布式的道路能见度检测与预警信息系统。具体而言,本文的研究内容包括:(1)系统总体框架设计。在分析我国交通行业能见度监测与预警需求的基础上,设计了基于监控视频分析的高速公路能见度检测与预警系统的整体框架。该框架主要由上位机和下位机两部分组成:下位机围绕高速公路现场的视频监控系统搭建,用于视频分析和能见度计算;上位机部署于路网信息中心,向用户提供能见度监控、低能见度预警处置、公共信息发布等功能。(2)能见度检测方法。该方法利用密集布设于高速公路沿线的监控摄像头,在摄像头视野内安装固定规格与颜色模式的参照物挡板,并基于经典能见度理论及视频图像分析技术,通过分析摄像头所采集的参照物图像的失真程度来计算实时能见度值。实验验证表明,该算法具有较高的检测精度,可以很好地满足我国高速公路的能见度检测应用需求。(3)能见度检测与预警系统研发。该系统的下位机子系统运行于现场嵌入式工控机,其核心软件程序具有插件式体系结构;上位机子系统基于JAVA EE技术研发,具有包含实体层、业务逻辑层、表现层等的分层体系结构。示范应用表明,该系统能够持续全面检测高速公路沿线能见度,并及时发布第能见度预警,具有重要的应用价值。
[Abstract]:In recent years, due to the deterioration of the natural environment, the impact of low visibility weather caused by haze, dust and strong light on the safe driving of expressways in China has become more and more serious.The demand of visibility weather detection and early warning along highway becomes more and more urgent.Compared with the conventional visibility measuring instruments, such as expensive cost, small sampling space, complex installation, difficult maintenance and other shortcomings, the traffic surveillance video recorded by the monitoring camera electromechanical equipment along the highway contains rich visibility information.It provides a new idea for highway visibility detection.In view of this, this paper designs a visibility detection method based on surveillance video, which is taken by the surveillance camera along the highway.A distributed road visibility detection and early warning information system is developed around this method.Specifically, the content of this paper includes the design of the general framework of the system.On the basis of analyzing the requirement of visibility monitoring and early warning in China's transportation industry, the whole frame of highway visibility detection and warning system based on surveillance video analysis is designed.The frame consists of two parts: the lower computer is built around the scene video surveillance system of the freeway, which is used for video analysis and visibility calculation; the upper computer is deployed in the network information center to provide visibility monitoring to users.Low visibility warning disposal, public information release and other functions. 2) visibility detection method.Based on the classical visibility theory and video image analysis technology, the method uses the surveillance camera which is located in the highway, and installs the reference baffle with fixed specifications and color patterns in the view of the camera.The real-time visibility is calculated by analyzing the distortion of the reference image collected by the camera.The experimental results show that the algorithm has high detection accuracy and can meet the requirement of visibility detection and warning system of freeway in China.The sub-system of the system runs in the field embedded industrial control computer, its core software program has the plug-in architecture, the upper computer subsystem is developed based on JAVA EE technology, it has the entity layer, the business logic layer, and the sub-system of the upper computer is based on the technology of JAVA EE.Hierarchical architecture representing layers, etc.The demonstration application shows that the system can continuously and comprehensively detect the visibility along the highway and issue the visibility warning in time. It has important application value.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U495;TP391.41

【相似文献】