基于视频的路侧停驶事件实时检测算法

详细信息    查看全文 | 推荐本文 |

英文篇名：Real-time Detection of Roadside Parking Events Based on Video 作者：胡煦 ; 黄俊 ; 袁梅 英文作者：Hu Xu;Huang Jun;Yuan Mei;School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications;Key Laboratory of Signal and Information Processing of Chongqing, Chongqing University of Posts and Telecommunications; 关键词：路侧停车管理 ; 目标检测 ; 多目标跟踪 ; 轨迹分析 ; 路侧停驶事件检测 英文关键词：Roadside Parking Managment;;Target Detection;;Multi-target Traking;;Trajectory analysis;;Parking Behavior Event Detection 中文刊名：HBYD 英文刊名：Information & Communications 机构：重庆邮电大学通信与信息工程学院;重庆邮电大学信号与信息处理重庆市重点实验室; 出版日期：2019-03-15 出版单位：信息通信 年：2019 期：No.195 基金：国家自然科学基金(61671095) 语种：中文; 页：HBYD201903002 页数：4 CN：03 ISSN：42-1739/TN 分类号：9-12

摘要

目前路侧停车收费多为人工巡视或车主自主缴费方式,停车管理的智能化及自动化程度较低,针对基于视频的路侧停车智能化管理方案,提出一种路侧停驶事件实时检测算法,提供单摄像头管理路侧多车位计时收费的能力。算法利用基于运动目标检测的多目标跟踪方法,实时提取车辆运动轨迹点。着重分析车辆路侧泊车过程,建立运动参数数学模型;提取轨迹点状态特征,建立停驶事件判别模型。实验表明,该算法能在车辆停进泊位及离开监控之前,预判车辆停驶行为,以便进行拍照取证;在车辆停入泊位及驶离监控区域之后,判定车辆停驶事件,用于停车计时。
        At present, most of the methods of roadside parking charges are manual charges or the owner pays consciously. The parking management method is less intelligent and less automated. Aiming at the video-based intelligent management scheme of roadside parking, a real-time detection algorithm of roadside parking events is proposed, which provides the ability of single camera to manage the multi-car parking time charging. The algorithm uses the multi-target tracking method based on moving target detection to extract vehicle trajectory points in real time. The paper focuses on analyzing the roadside parking process, establishes the mathematical model of the motion parameters, extracts the state characteristics of the trajectory points, and establishes the discriminating model of the parking events. Experiments show that the algorithm can predict the vehicle parking behavior before entering parking space and leaving the monitoring area, so as to take photos and obtain evidence; after entering parking space and leaving the monitoring area, it can determine the parking events for timing.

引文

[1]胡列格,郭添添.国内外城市停车发展策略思路研究[J].中国交通信息产业,2009(01):131-133.
    [2]Ivanov I.,Dufaux F.,Ha T.M.and Ebrahimi T.Towards generic detection of unusual events in video surveillance[C]//2009 Sixth IEEE International Conference on Advanced Video and Signal Based Surveillance,Genova,2009,pp.61-66.
    [3]Liu Chunmei,Hu Changbo,Liu Qingshan,et a1.Video event description in scene context[J].Neurocomputing,2013,119:82-93.
    [4]Ahmed Sk.A,Dogra D P,Kar Samarjit,et a1.Surveillance scene representation and trajectory abnormality detection using aggregation of multiple concepts[J].Expert Systems with Applications,2018,101:43-55.
    [5]Wang Xuan,Song Huansheng,Cui Hua.Pedestrian abnormal event detection based on multi-feature fusion in traffic video[J].Optik,2018,154:22-32.
    [6]于青青,等.基于视频跟踪轨迹的全过程路侧停车行为检测与识别技术[J].计算机与现化,2017(09):67-73.
    [7]蒋恩源,王学军.基于跟踪轨迹的车辆异常行为检测[J].吉林大学学报(信息科学版),2016,34(01):98-103
    [8]赵有婷,李熙莹,罗东华.基于视频车辆轨迹模型的交通事件自动检测方法研究[J].中山大学学报(自然科学版),2011,50(04):56-60+64.
    [9]肖进胜,刘婷婷,等.一种基于历史背景的混合高斯背景建模算法[J].湖南大学学报(自然科学版),2015,42(10):127-132.
    [10]Domadiya P,Shah P,Mitra S K.Fast and accurate foreground background separation for video surveillance[J].Lecture Notes in Computer Science,2015,9124:96-103.
    [11]高冬冬.基于车辆跟踪轨迹的停车和逆行检测研究[D].长安大学,2015.
    [12]刘畅,杨锁昌,汪连栋.粒子滤波理论在单目标跟踪中的应用综述[J].飞航导弹,2017(10):67-71+90.