基于监控视频的公路能见度检测方法
|
摘要
雾霾、沙尘等低能见度天气状况的频繁发生,严重影响到日常交通管理及出行者的生命、财产安全。针对低能见度环境下极易发生公路交通事故的问题,提出一种利用公路沿线已经架设的监控设备对各路段的监控视频数据进行实时图像处理分析的公路能见度检测方法。该方法是在传统暗原色先验算法的基础上,将监控视频帧转换为灰度图像,并结合自适应导向滤波算法和四叉树图像分割算法,提取监控图像所包含的空气透射率及大气光值等特征参数。然后基于能见度计算模型,依据监控设备与检测目标物之间的实际距离及偏移角度,估测道路的能见距离。实验结果表明,优化后的图像处理方法可以解决图像清晰度低、色彩失真、场景适应能力差等问题,且检测的能见度值误差率满足我国公路能见度检测精度要求。最后,结合能见度检测方法,提出了相应的能见度预警策略并应用于公路能见度检测与预警系统,为驾驶员实时提供雾区路段的能见度值信息和预警建议。
Low-visibility weathers, such as smog and sand storm, were causing tremendous risks to road traffic and travel, seriously affecting the daily traffic management and the safety of life and property of travelers. Aiming at the problem that highway was prone to traffic accidents in low visibility environment, a highway visibility detection method was proposed based on real-time image processing and analysis of monitoring video data of each section using monitoring equipment installed along the highway. Based on the traditional dark channel prior, the highway visibility detection method converted the surveillance video frame into gray image, and extracted the air transmittance and atmospheric light value contained in the surveillance image by the adaptive steering filter algorithm and quadtree image segmentation algorithm. Then based on the visibility calculation model, the visibility distance of the road was estimated according to the actual distance and offset angle between the monitoring equipment and the detection object. The experimental results showed that the optimized image processing method could solve the problems of low image clarity, color distortion, and poor scene adaptability.The detection of visibility error rate meets the accuracy requirements of highway visibility detection in China. Finally, based on the visibility detection method, the corresponding visibility warning strategies were applied in a road visibility detection and early warning system to provide the driver with real-time visibility values and warning recommendations for the fog areas.
Low-visibility weathers, such as smog and sand storm, were causing tremendous risks to road traffic and travel, seriously affecting the daily traffic management and the safety of life and property of travelers. Aiming at the problem that highway was prone to traffic accidents in low visibility environment, a highway visibility detection method was proposed based on real-time image processing and analysis of monitoring video data of each section using monitoring equipment installed along the highway. Based on the traditional dark channel prior, the highway visibility detection method converted the surveillance video frame into gray image, and extracted the air transmittance and atmospheric light value contained in the surveillance image by the adaptive steering filter algorithm and quadtree image segmentation algorithm. Then based on the visibility calculation model, the visibility distance of the road was estimated according to the actual distance and offset angle between the monitoring equipment and the detection object. The experimental results showed that the optimized image processing method could solve the problems of low image clarity, color distortion, and poor scene adaptability.The detection of visibility error rate meets the accuracy requirements of highway visibility detection in China. Finally, based on the visibility detection method, the corresponding visibility warning strategies were applied in a road visibility detection and early warning system to provide the driver with real-time visibility values and warning recommendations for the fog areas.
引文
[1]李晓雷,唐伯明,曾超.基于突变论的高速公路团雾驾驶行为瞬态分析[J].中国公路学报,2018,31(4):78-87.
[2]朱岩,张洋,张丽红,等.低能见度下高速公路主动诱导技术的应用[J].中国交通信息化,2016(6):83-85.
[3]周晓旭.低能见度高速公路主动发光诱导系统设计[J].山西交通科技,2017(4):119-121.
[4]CHEN Z Z.PTZ Visibility Detection Based on Image Luminance Changing Tendency[C]//Proceedings of 2016 International Conference on Optoelectronics and Image Processing(ICOIP).Warsaw:IEEE,2016:15-19.
[5]HU X C,ZHUO L,LI X G.A Moving Objects Based Real-Time Defogging Method for Traffic Monitoring Videos[C]//Proceedings of 2014 19th International Conference on Digital Signal Processing(DSP).Hong Kong:IEEE,2014:1-6.
[6]KWON T M.Atmospheric Visibility Measurements Using Video Cameras:Relative Visibility[R].Duluth,Americas:University of Minnesota Duluth,2004.
[7]GUO F,PENG H,TANG J,et al.Visibility Detection Approach to Road Scene Foggy Images[J].KSⅡTransactions on Internet&Information Systems,2016,10(9):4419-4441.
[8]韩静,于平平,王震洲,等.基于相机响应曲线拟合的道路能见度检测算法[J].燕山大学学报,2018,42(3):272-277.
[9]WU H T,LAN J H.A Novel Fog-Degraded Image Restoration Model of Golden Scale Extraction in Color Space[J].Arabian Journal for Science and Engineering,2018,43(12):6801-6821.
[10]HE K M,SUN J,TANG X O.Single Image Haze Removal Using Dark Channel Prior[J].IEEE Transaction on Pattern Analysis Machine Intelligence,2011,33(12):2341-2353.
[11]ZHANG C L,SUN T,CHEN J R,et al.A Visibility Monitoring System Utilizing Roadside Video Camera Facilities for Highway Systems[C]//Proceedings of 2017 3rd International Conference on Information Management(ICIM).Chengdu:IEEE,2017:486-490.
[12]黄希,王杰,王恒,等.一种基于Arduino和BH1750的前向散射式能见度检测装置设计[J].南通大学学报(自然科学版),2015,14(4):6-10.
[13]曾浩,尚媛园,丁辉,等.基于暗原色先验的图像快速去雾[J].中国图象图形学报,2015,20(7):914-921.
[14]FU F Z,LIU F.Wavelet-Based Retinex Algorithm for Unmanned Aerial Vehicle Image Defogging[C]//Proceedings of 2015 8th International Symposium on Computational Intelligence and Design(ISCID).Hangzhou:IEEE,2015:426-430.
[15]FATTAL R.Single image dehazing[J].ACM Transactions on Graphics,2008,27(3):72.
[16]江苏省气象科学研究所.高速公路能见度监测及浓雾的预警预报:QX/T 76-2007[S].北京:气象出版社,2007.
[17]赵瑞,胡辽林,赵江涛.基于暗原色先验理论的大气能见度测量[J].激光与光电子学进展,2016(6):72-79.
[18]董沛武,杨玫,刘世国.干线公路安全预警可拓物元评价模型及策略研究[J].中国软科学,2014(11):140-150.
[2]朱岩,张洋,张丽红,等.低能见度下高速公路主动诱导技术的应用[J].中国交通信息化,2016(6):83-85.
[3]周晓旭.低能见度高速公路主动发光诱导系统设计[J].山西交通科技,2017(4):119-121.
[4]CHEN Z Z.PTZ Visibility Detection Based on Image Luminance Changing Tendency[C]//Proceedings of 2016 International Conference on Optoelectronics and Image Processing(ICOIP).Warsaw:IEEE,2016:15-19.
[5]HU X C,ZHUO L,LI X G.A Moving Objects Based Real-Time Defogging Method for Traffic Monitoring Videos[C]//Proceedings of 2014 19th International Conference on Digital Signal Processing(DSP).Hong Kong:IEEE,2014:1-6.
[6]KWON T M.Atmospheric Visibility Measurements Using Video Cameras:Relative Visibility[R].Duluth,Americas:University of Minnesota Duluth,2004.
[7]GUO F,PENG H,TANG J,et al.Visibility Detection Approach to Road Scene Foggy Images[J].KSⅡTransactions on Internet&Information Systems,2016,10(9):4419-4441.
[8]韩静,于平平,王震洲,等.基于相机响应曲线拟合的道路能见度检测算法[J].燕山大学学报,2018,42(3):272-277.
[9]WU H T,LAN J H.A Novel Fog-Degraded Image Restoration Model of Golden Scale Extraction in Color Space[J].Arabian Journal for Science and Engineering,2018,43(12):6801-6821.
[10]HE K M,SUN J,TANG X O.Single Image Haze Removal Using Dark Channel Prior[J].IEEE Transaction on Pattern Analysis Machine Intelligence,2011,33(12):2341-2353.
[11]ZHANG C L,SUN T,CHEN J R,et al.A Visibility Monitoring System Utilizing Roadside Video Camera Facilities for Highway Systems[C]//Proceedings of 2017 3rd International Conference on Information Management(ICIM).Chengdu:IEEE,2017:486-490.
[12]黄希,王杰,王恒,等.一种基于Arduino和BH1750的前向散射式能见度检测装置设计[J].南通大学学报(自然科学版),2015,14(4):6-10.
[13]曾浩,尚媛园,丁辉,等.基于暗原色先验的图像快速去雾[J].中国图象图形学报,2015,20(7):914-921.
[14]FU F Z,LIU F.Wavelet-Based Retinex Algorithm for Unmanned Aerial Vehicle Image Defogging[C]//Proceedings of 2015 8th International Symposium on Computational Intelligence and Design(ISCID).Hangzhou:IEEE,2015:426-430.
[15]FATTAL R.Single image dehazing[J].ACM Transactions on Graphics,2008,27(3):72.
[16]江苏省气象科学研究所.高速公路能见度监测及浓雾的预警预报:QX/T 76-2007[S].北京:气象出版社,2007.
[17]赵瑞,胡辽林,赵江涛.基于暗原色先验理论的大气能见度测量[J].激光与光电子学进展,2016(6):72-79.
[18]董沛武,杨玫,刘世国.干线公路安全预警可拓物元评价模型及策略研究[J].中国软科学,2014(11):140-150.