道路视觉运行状态评测系统研究
|
摘要
最近几年来,我国的高速公路路网日趋成熟,高速公路的建设也飞速的发展,高速公路的监管和运营方面面临更高的要求和挑战,本论文通过对高速公路路段的交通流以运行特征进行分析研究,对高速公路的运行状态进行评测分析,以便给高速公路的监控管理工作提供更科学可靠的依据。从而更方便有效的开展高速路网的交通调度工作,充分发挥高速公路路网的整体效益,提高路网运营管理水平和服务水平。本论文结合我国高速公路现状,从运行状况的指标体系建立、交通事件检测子系统、交通流量参数——占空比检测三个方面,对高速公路路段运行状况评价分析进行了研究。这三个方面相互联系,从不同侧面反映了路段的实时运行状况。
本文基于数字图像处理和计算机视觉技术,对系统中的关键技术进行了深入研究。在系统设计方面,考虑到当前交通领域现有监控设备,设计了与之吻合的“视频图像采集模块+视频检测服务器+视频检测终端”模式。
在运行状况动态综合评价研究方面,本论文根据我国高速公路的普遍现状建立综合评价指标体系；在分析现有综合评价方法的基础上,提出了适合路段运行状况的评价方法模型。该方法采用先状态分类后趋势评价的两步评价方式,将适用于时间序列的动态综合评价方法应用于路段运行状况的评价,该方法能获得比较准确的运行状态分类和运行趋势分析,适用于路段运行状况的动态实时评价。
在算法的研究过程中,通过对目标检测子系统中的背景模型算法、阴影检测算法、运动目标检测算法以及运动目标跟踪算法的分析研究提出了车辆类型识别的技术：基于几何特征分类器设计算法和基于不变矩分类器设计算法,为占空比的检测提供了良好的基础。
In recent years,along with the rapid development of freeway in our country,the freeway network is developing gradually.At the same time,freeway management has to be face with higher requirement and greater challenge than before.Through the study on comprehensive evaluation and analysis of operation status of freeway section,the paper provided the scientific bases and datum for freeway governor,and to improve the level of freeway network management and service,further.Considering the actuality of freeway in our country,this dissertation is studied for three important aspects on freeway section:dynamic comprehensive evaluation of operation status,the child system of incident detection and traffic flow parameters-- occupancy space. The three aspects are related to one another,and each aspect reflects the dynamic operation status of freeway section from a special side.
Based on the digital image processing and the computer vision technology, the paper has conducted in-depth research in the key technologies of the system. In the aspect of system designing, considered the existing monitoring equipments in current transportation fields, the paper has designed the compatibly mode of "the video image gathering module + video detection server + video detection terminal".
In the first aspect,this dissertation firstly presents the evaluation attribute system based on analyzing common attributes,then the dissertation study and analysis all existing comprehensive evaluation methods.According to the requirement of dynamic comprehensive evaluation of operation status on freeway section,the dissertation presents a new dynamic evaluation method called"classify status first,evaluate trend later".It is the first time that the dynamic comprehensive evaluation method is applied on system evaluation of traffic field.The application example proves the validity of the new method.
In the study of algorithms,the paper brought forward the technology of vehicle type recognition,such as classification algorithms based on Geometrical characteristics and classification algorithms based on constant moment,thought the algorithm analysis of background model and shadow detection and moving object detection and moving target tracking.All above provided a good basis for occupancy space testing.
本文基于数字图像处理和计算机视觉技术,对系统中的关键技术进行了深入研究。在系统设计方面,考虑到当前交通领域现有监控设备,设计了与之吻合的“视频图像采集模块+视频检测服务器+视频检测终端”模式。
在运行状况动态综合评价研究方面,本论文根据我国高速公路的普遍现状建立综合评价指标体系；在分析现有综合评价方法的基础上,提出了适合路段运行状况的评价方法模型。该方法采用先状态分类后趋势评价的两步评价方式,将适用于时间序列的动态综合评价方法应用于路段运行状况的评价,该方法能获得比较准确的运行状态分类和运行趋势分析,适用于路段运行状况的动态实时评价。
在算法的研究过程中,通过对目标检测子系统中的背景模型算法、阴影检测算法、运动目标检测算法以及运动目标跟踪算法的分析研究提出了车辆类型识别的技术：基于几何特征分类器设计算法和基于不变矩分类器设计算法,为占空比的检测提供了良好的基础。
In recent years,along with the rapid development of freeway in our country,the freeway network is developing gradually.At the same time,freeway management has to be face with higher requirement and greater challenge than before.Through the study on comprehensive evaluation and analysis of operation status of freeway section,the paper provided the scientific bases and datum for freeway governor,and to improve the level of freeway network management and service,further.Considering the actuality of freeway in our country,this dissertation is studied for three important aspects on freeway section:dynamic comprehensive evaluation of operation status,the child system of incident detection and traffic flow parameters-- occupancy space. The three aspects are related to one another,and each aspect reflects the dynamic operation status of freeway section from a special side.
Based on the digital image processing and the computer vision technology, the paper has conducted in-depth research in the key technologies of the system. In the aspect of system designing, considered the existing monitoring equipments in current transportation fields, the paper has designed the compatibly mode of "the video image gathering module + video detection server + video detection terminal".
In the first aspect,this dissertation firstly presents the evaluation attribute system based on analyzing common attributes,then the dissertation study and analysis all existing comprehensive evaluation methods.According to the requirement of dynamic comprehensive evaluation of operation status on freeway section,the dissertation presents a new dynamic evaluation method called"classify status first,evaluate trend later".It is the first time that the dynamic comprehensive evaluation method is applied on system evaluation of traffic field.The application example proves the validity of the new method.
In the study of algorithms,the paper brought forward the technology of vehicle type recognition,such as classification algorithms based on Geometrical characteristics and classification algorithms based on constant moment,thought the algorithm analysis of background model and shadow detection and moving object detection and moving target tracking.All above provided a good basis for occupancy space testing.
引文
[1]王春波.交通监控中视觉信息的检测与跟踪技术的研究[D].上海：上海交通大学,2000
[2]金会庆,戴平,张树林.智能运输系统(ITS)研究现状及展望[J].人类工效学,2001,7(3)：39-41
[3]何刚.高速公路交通运行分析与评价[C].常州：中国公路学会交通工程技术及装备发展研讨会2000.9：65-69
[4]明平顺.高速公路运行状况分析与对策[J].中南汽车运输.1997(3)：35-37
[5]周刚,张国良等.高速公路交通运行特性分析研究[J].广东公路科技.2001(9)：5-8
[6]Chao Chen.Freeway Performance Measurement System (PeMS) [D]:[博士学位论文].University of California,Berkeley,2002.9-52
[7]王炜,过秀成.交通工程学[M].南京：东南大学出版社,2000：339-342
[8]曹仁川,刘天宇,董铁军.全视频技术在电子警察系统中的应用[J].中国信息产业,2004,104-105
[9]陈红艳.交通流视频检测中的摄像机标定技术研究[D].吉林：吉林大学,2007
[10]PREVEDOUROS P D, JI Xiaojin, PAOANDEOU K. Video incidentdetection tests in freeway tunnels [EB/OL], http://pubsindex.trb.org/document/view/default.asp?lbid=776527,2009-2-25
[11]JTG B01-2003,公路工程技术标准[S].北京：中华人民共和国交通部,2004
[12]杨枝灵,王开.Visual C++数字图像获取处理及实践应用[M].北京：人民邮电出版社,2003
[13]常成利,周刚.高速公路路段通行能力分析方法的探索与实践[J].公路交通科技.2003,20(2)：68-72
[14]孙大跃.高速公路紧急救援体系研究[D].西安：长安大学,2006
[15]MittalA, HuttenlocherD. Scene modeling forwide area surveillance and image synthesis [A]. In: Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition [C]. Hilton Head Island, SC, USA,2000:160-167
[16]Wren C R, Azarbayejani A, Darrell T, et al. Pfinder:real-time tracking of the human body [J]. IEEETransactions on Pattern Analysis and Machine Intel-ligence,1997,19(7):780-785
[17]Joonho Ko and Randall L Guensler. Characterization of Congestion Based on Speed Distribution:A Statistical Approach Using Gaussian Mixture Model. Transportation Research Board,2005
[18]王炜,陈学武,陆建.城市交通系统可持续发展理论体系研究[M].北京：科学出版社,2003
[19]杨佩昆,吴兵.交通管理与控制[M].(第二版).北京：人民交通出版社,2003.153-157
[20]陆化普,王建伟,李江平等.城市交通管理评价体系[M].北京：人民交通出版社,2003.72-80
[21]杨兆升.关于智能运输系统的关键理论——综合路段行程时间预测的研究[J].交通运输工程学报.2001,1(1)：65-68
[22][42]邓毅萍.高速公路路段运行状况评价与分析研究：[硕士学位论文].南京：东南大学硕士学位论文,2005
[23]王圣男,郁梅,蒋刚毅.基于视频图像处理的车辆检测与跟踪方法综述[J].计算机应用研究,2005,22(9)：9-14
[24]R Cucchiara, C Grana, M Piccardi et al. Detecting Objects, Shadow and Ghosts in Video Streams by Exploiting Color and Motion Information[C]. In:Proceedings of 11th International Conference on Image Analysis and Processing(ICIAP 2001),2001
[25]B. horn, B. Schunck. Determining optical flow [J]. Artificial Intelligence,1981,17:185-203
[26]王栓,艾海舟,何克忠.基于差分图像的多运动目标的检测与跟踪[J].中国图像图形学报,1999,4(6)：470-475
[27]Naoya Ohta. A statistical approach to background subtraction for surveillance systems [A]. In: Proceeding of the International Conference on Computer Vision [C], Vancouver, Canada,2001,2:481-486
[28]姚红革,耿军雪.基于卡尔曼预测的视频目标实时跟踪[J].西安工业大学学报,2007,27(2)：171-175
[29]Richard O D,Peter E H,David G S.著.李宏东,姚天翔等译.模式分类(原第二版)[M]北京：机械工业出版社,2003.
[30]张宏林著.’Visual C++数字图像模式识别技术及工程实践[M].北京：人民邮电出版社,2003.
[31]He J,Tan AH,Tan CL.A comparative study on Chinese text categorization methods.In:Proceedings of the International Workshop on Text and Web Mining.Singapore:Melboume,2000.24-35.
[32]Cover TM,Hart PE.Nearest neighbor pattern classification.IEEE Transactions onInformation Theory,1968,IT1-3:21-27.
[33]Li RL,Hu YF.Noise reduction to text categorization based on density for kNN.In:Proceeding of the Second International Conference on Machine Learning and Cybernetics.Xi'an,2003.3119-3124.
[34]翟乃斌,苏建,宋年秀,等.基于不变矩的计算机视觉车辆类型识别技术研究[J].武汉理工大学学报(信息与管理版),2007.4(4)
[2]金会庆,戴平,张树林.智能运输系统(ITS)研究现状及展望[J].人类工效学,2001,7(3)：39-41
[3]何刚.高速公路交通运行分析与评价[C].常州：中国公路学会交通工程技术及装备发展研讨会2000.9：65-69
[4]明平顺.高速公路运行状况分析与对策[J].中南汽车运输.1997(3)：35-37
[5]周刚,张国良等.高速公路交通运行特性分析研究[J].广东公路科技.2001(9)：5-8
[6]Chao Chen.Freeway Performance Measurement System (PeMS) [D]:[博士学位论文].University of California,Berkeley,2002.9-52
[7]王炜,过秀成.交通工程学[M].南京：东南大学出版社,2000：339-342
[8]曹仁川,刘天宇,董铁军.全视频技术在电子警察系统中的应用[J].中国信息产业,2004,104-105
[9]陈红艳.交通流视频检测中的摄像机标定技术研究[D].吉林：吉林大学,2007
[10]PREVEDOUROS P D, JI Xiaojin, PAOANDEOU K. Video incidentdetection tests in freeway tunnels [EB/OL], http://pubsindex.trb.org/document/view/default.asp?lbid=776527,2009-2-25
[11]JTG B01-2003,公路工程技术标准[S].北京：中华人民共和国交通部,2004
[12]杨枝灵,王开.Visual C++数字图像获取处理及实践应用[M].北京：人民邮电出版社,2003
[13]常成利,周刚.高速公路路段通行能力分析方法的探索与实践[J].公路交通科技.2003,20(2)：68-72
[14]孙大跃.高速公路紧急救援体系研究[D].西安：长安大学,2006
[15]MittalA, HuttenlocherD. Scene modeling forwide area surveillance and image synthesis [A]. In: Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition [C]. Hilton Head Island, SC, USA,2000:160-167
[16]Wren C R, Azarbayejani A, Darrell T, et al. Pfinder:real-time tracking of the human body [J]. IEEETransactions on Pattern Analysis and Machine Intel-ligence,1997,19(7):780-785
[17]Joonho Ko and Randall L Guensler. Characterization of Congestion Based on Speed Distribution:A Statistical Approach Using Gaussian Mixture Model. Transportation Research Board,2005
[18]王炜,陈学武,陆建.城市交通系统可持续发展理论体系研究[M].北京：科学出版社,2003
[19]杨佩昆,吴兵.交通管理与控制[M].(第二版).北京：人民交通出版社,2003.153-157
[20]陆化普,王建伟,李江平等.城市交通管理评价体系[M].北京：人民交通出版社,2003.72-80
[21]杨兆升.关于智能运输系统的关键理论——综合路段行程时间预测的研究[J].交通运输工程学报.2001,1(1)：65-68
[22][42]邓毅萍.高速公路路段运行状况评价与分析研究：[硕士学位论文].南京：东南大学硕士学位论文,2005
[23]王圣男,郁梅,蒋刚毅.基于视频图像处理的车辆检测与跟踪方法综述[J].计算机应用研究,2005,22(9)：9-14
[24]R Cucchiara, C Grana, M Piccardi et al. Detecting Objects, Shadow and Ghosts in Video Streams by Exploiting Color and Motion Information[C]. In:Proceedings of 11th International Conference on Image Analysis and Processing(ICIAP 2001),2001
[25]B. horn, B. Schunck. Determining optical flow [J]. Artificial Intelligence,1981,17:185-203
[26]王栓,艾海舟,何克忠.基于差分图像的多运动目标的检测与跟踪[J].中国图像图形学报,1999,4(6)：470-475
[27]Naoya Ohta. A statistical approach to background subtraction for surveillance systems [A]. In: Proceeding of the International Conference on Computer Vision [C], Vancouver, Canada,2001,2:481-486
[28]姚红革,耿军雪.基于卡尔曼预测的视频目标实时跟踪[J].西安工业大学学报,2007,27(2)：171-175
[29]Richard O D,Peter E H,David G S.著.李宏东,姚天翔等译.模式分类(原第二版)[M]北京：机械工业出版社,2003.
[30]张宏林著.’Visual C++数字图像模式识别技术及工程实践[M].北京：人民邮电出版社,2003.
[31]He J,Tan AH,Tan CL.A comparative study on Chinese text categorization methods.In:Proceedings of the International Workshop on Text and Web Mining.Singapore:Melboume,2000.24-35.
[32]Cover TM,Hart PE.Nearest neighbor pattern classification.IEEE Transactions onInformation Theory,1968,IT1-3:21-27.
[33]Li RL,Hu YF.Noise reduction to text categorization based on density for kNN.In:Proceeding of the Second International Conference on Machine Learning and Cybernetics.Xi'an,2003.3119-3124.
[34]翟乃斌,苏建,宋年秀,等.基于不变矩的计算机视觉车辆类型识别技术研究[J].武汉理工大学学报(信息与管理版),2007.4(4)