基于神经网络的汽车车型识别系统论文下载DOC

I摘要近年来，随着我国国民经济的不断发展，交通运输越来越繁忙，对交通管理提出了新的要求。在过桥收费站、大型停车场、城市道路监管、治安卡口、港口和飞机场等实际交通系统中，需要对汽车车型进行识别，以便收取相应的费用和提高交通系统车辆监控和自动化程度。因此，如何对上述各类车辆收费站实现现代化的管理，具有重要的现实意义。针对基于神经网络的汽车车型识别系统中的识别技术问题，本文从以下四个部分进行了研究和探讨：第一部分，论述了车型识别的研究背景和意义，详细分析了我国目前车型识别系统的应用现状及研究现状，指出了目前国内外应用系统及其车型识别方法存在的缺陷与不足。第二部分，提出了车型识别的模型。首先对采集的车辆图像进行预处理，通过灰度转换、图像平滑等方法剔除噪音，以提高图像质量。然后对其进行分割并提取特征，在这个过程中经过图像的二值化处理]16[，拉普拉斯边缘检测、图像横向填充与纵向填充、轮廓提取、图像修正，再提取出图像车型的上顶长、高、前底长、后底长等特征参数。结合所提取的特征参数进行车型识别。第三部分，设计拉普拉斯边缘检测算子的汽车识别算法。采用序列差影法进行背景剔除；边缘检测之后的图像进行离散噪点的剔除。采用轮廓法对横/纵向填充图像进行轮廓提取。第四部分，设计实现一个车型识别系统，以此检验论文理论研究的可行性，并通过不断地实际测验来改良算法。本文以VC++6.0为软件平台，以Matlab为仿真平台，编程实现了基于序列图像的车型识别系统，通过实验数据分析表明，本文给出的识别方法能得到较好的识别结果。关键词：智能交通，车型识别，图像处理，特征提取IIABSTRACTWiththedevelopmentofournationaleconomy,traffichasbecomeheavierinrecentyears,whichhasputforwardnewrequirementsfortrafficmanagement.Inactualtrafficsystemsuchasbridgetollstation,largeparkinglot,portsandairports,vehiclerecognitionisnecessarytochargecorrespondingfeesandimprovetheautomationandvehiclemonitoringofthetrafficsystem.Therefore,themodernmanagementofthetollstationsabovehasimportantpracticalsignificance.Facingthetechnicalproblemsofthevehiclerecognitionsystembasedontheneuralnetwork,thispaperconductsresearchanddiscussionfromthefollowingfourparts:Inthefirstpart,wediscussthebackgroundandsignificanceofthevehiclerecognition,thendetailedanalyzesthepresentsituationoftheapplicationofvehiclerecognitionsystemandthepresentresearchstatus,andpointouttheshortcomingsofvehiclerecognitionsystemsathomeandabroad.Inthesecondpart,weputforwardthemodelofthevehiclerecognitionsystem.Firstweconductthepreprocessingoftheobtainedvehicleimage,toreducethenoiseoftheimagebygrayingandsmoothingit,toimprovethequalityoftheimage.Nextstepisthesegmentationandfeatureextracting.Aftertheimagebinarizationprocessing,Laplaceedgedetection,imagehorizontalandverticalfilling,contourextraction,imagecorrection,weextractthelengthandheightofthefrontandbackofthecar.Vehiclerecognitioncanbedonewiththecharacteristicparametersextracted.Inthethirdpart,wedesignavehiclerecognitionalgorithmusingLaplaceedgedetectionoperator,andeliminatebackgroundusingsequencedifferenceimagemethod.Theneliminatethediscretenoisepointsafteredgedetectionoftheimage.,andextractthecontourofthehorizontal/verticalfilledimageusingcontourmethod.Inthefourthpart,wedesignedavehiclerecognitionsystemtotestthefeasibilityofthetheory,andimprovethealgorithmthroughthecontinuousactualtestIIIBasedonsoftwareplatformofVC++6.0,withMatlabsimulationplatform,wefinishthevehiclerecognitionsystembasedonimagesequences.Throughtheexperimentaldataanalysisshowed,wecanfindthatthismethodofidentificationpresentedcangetgoodrecognitionresults.Keywords：intelligenttransportation,vehiclerecognition,imageprocessing,featureextractionIV目录前言...............................................................11绪论..............................................................31.1选题的目的及意义.............................................31.2车型识别的研究现状...........................................41.3论文研究的思路与流程.........................................41.4车型图像处理平台.............................................51.4.1车型图像硬件平台.......................................51.4.2车型图像软件平台.......................................51.4.3车型图像仿真平台.......................................62车辆图像的分离与处理..............................................82.1车辆图像的分离...............................................82.2图像处理的基本知识..........................................112.2.1BMP图像的读取........................................112.2.2图像的灰度化处理......................................112.2.3图像的平滑处理........................................122.2.4灰度图像的二值化处理..................................132.2.5图像的边缘化检测......................................132.3车辆图像的处理结果..........................................162.3.1车辆BMP图像显示与读取................................162.3.2车辆图像灰度化........................................192.3.3车辆图像平滑处理——中值滤波..........................202.3.4车辆图像的二值化及二值反向............................222.3.5车辆图像的边缘检测....................................243车辆的特征提取...................................................263.1特征提取....................................................263.2车型轮廓提取................................................263.3目标区域的填充..............................................29V3.3.1图像横向填充..........................................293.3.2图像的纵向填充........................................314车型识别.........................................................324.1车辆轮廓识别................................................324.2车辆特征参数的提取与计算....................................324.3车型特征匹配................................................335总结与展望.......................................................355.1总结与分析..................................................355.2未来展望....................................................35致谢..............................................................37参考文献...........................................................38--1前言随着全球经济的告诉发展，社会对交通运输的需求持续增长，无论是发达国家还是发展中国家都面临着城市交通拥堵、交通事故频发、交通环境恶化以及能源短缺等问题。智能交通系统(IntelligentTranspo