欢迎来到天天文库
浏览记录
ID:36514932
大小:1.34 MB
页数:50页
时间:2019-05-11
《支持向量机在视频交通信息检测系统中的应用研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、Y87013S支持向量机在视频交通信息检测系统中的应用研究StudyofSupportVectorMachineanditsApplicationsinVideoTrafficInformationDetectionSystem作者姓名一.至亘丞学位类型堂压题±学科、专业盐簋扭鏊佳兰堡途研究方向数握度生垒攫住导师及职称王浩兹援2006年4月10日StudyofSupportVectorMachineanditsApplicationsinVideoTrafficInformationDetectionSystemAbstractStatisticalLearningTheor
2、y(SLT)isalearningtheorywhichspecializesinmachinelearningwithfinitesamples.AsalearningmethodbasedonSLT,SupportVectorMachine(SVM)hastheadvantagesofglobalsolutions,welladaptively.highgeneralizationabilityandmaturityintheory.SLTandSVMarethehot.spotinthefieldofmachinelearningnowadays.Withthedeve
3、lopmentofurbantrafficmanagement,thedetectionoftrafileinformationiSpaidmoreandmoreattention.VideodetectioniSsuperiortoothermethods.VideoTrafficInformationDetectionSystemiSacomputerprocessingsystemusingimageprocessingtechnologiesandaimstogathertrafficinformationsuchaspassedvehiclecount,vehicl
4、espeedandvehicletypes.Inthisthesis。weintroduceSVMtothesystemofVideoTrafficInformationDetection;themainworkiSdescribedasf01lows:f1、WesummarizethelatestresearchachievementsanddevelopmentofSLl’:presenttheconceptionsofSU’andtheprinciplesofSVM;r2、IntroducethesuperiorityofvideotraffiCinformationd
5、etectiontechnologies.presentthetheoriesandmethodsthatareUsedinaVideoTrafficInformationDetectionSystem;f31Accomplishasystemofvideotrafficinformationdetectionusingcurrentmethodandanalyzeitsfunctionandshortness.r41ApplythetheoryofSVMtoVideoTrafficInformationDeteetionSystem.Throughclassifyingth
6、etrafficflOWimagesindetectionareaandtracesvehiclestogettheinformationofpassedvehiclecount,vehiclespeedandvehicletypes.Lastly,wetakeacomparisonbetweenSVMmethodandcurrentmethod.Keywords:Statisticallearningtheory,Supportvectormachine,Videotrafficinformationdetectionsystem插图清单图2.1机器学习的基本模型⋯⋯⋯⋯⋯
7、⋯⋯⋯..图2.2VC维示意图⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯..图2.2结构风险最小化示意图⋯⋯⋯⋯⋯⋯⋯⋯图3.1最优超平面示意图⋯⋯⋯⋯⋯⋯⋯⋯⋯图3.2支持向量机示意图⋯⋯⋯⋯⋯⋯⋯⋯⋯图3.3支持向量机用于二维样本分类的例子..图3.4多类识别决策树⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯.图4.1系统基本上作流程⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯图4.2道路交通视频图像⋯⋯⋯⋯⋯⋯⋯⋯⋯.图4.3图像预处理⋯⋯⋯⋯⋯⋯⋯⋯⋯.⋯⋯⋯..图4.4车辆经过检测区域的过程⋯⋯⋯⋯⋯⋯⋯图4.5多值分类决策树⋯⋯⋯.⋯⋯⋯..⋯⋯⋯⋯图4.6
此文档下载收益归作者所有