西南交通大学学报2018，Vol.53Issue(1)：136-145,10.DOI:10.3969/j.issn.0258-2724.2018.01.017
基于监控视频的高速公路能见度检测方法及系统
Monitoring Video-Based Visibility Detection Method and System for Highways in China
摘要
Abstract
In view of the urgent demands for visibility detection,alarm and emergency disposal on highways in China,a monitoring video-based method for highway visibility detection was proposed. In this method,images of fixed-distance reference objects were first extracted from monitoring camera images,and then candidate visibility values were calculated based on greyscale image distortion analysis on these images. The final visibility result was computed by applying data fusion on the candidate visibility values. Based on this method,a visibility detection,alarm and emergency disposal system was developed. By utilizing the calculated visibility values,this system provides support for intelligent transportation tasks,such as visibility alarm,emergency disposal,traffic control,and information publication,both locally and globally. Experimental results show that the average detection error of this method is 13. 4％,thus meeting the accuracy and performance requirements of highway visibility detection. Moreover,this method and the corresponding system can fully reuse the monitoring camera electromechanical systems distributed along the highways;therefore they are suitable for application in operation management and efficiency improvement works on highways in China.关键词
交通工程/高速公路/能见度检测/图像处理/数据融合Key words
transportation engineering/highway/visibility detection/image processing/data fusion分类
交通工程引用本文复制引用
张昌利,孟颖,李鹏超,孙婷..基于监控视频的高速公路能见度检测方法及系统[J].西南交通大学学报,2018,53(1):136-145,10.基金项目
陕西省科技攻关项目(2015GY052) （2015GY052）
陕西省国际科技合作计划资助项目(2017-KW-015) （2017-KW-015）
国家留学基金委青年骨干教师出国研修项目(2016-3099)西安市科技计划资助项目(CXY1125-9). （2016-3099）
