基于形态学和神经网络的车牌识别算法（the license plate recognition algorithm based on neural network and morphology）

《基于形态学和神经网络的车牌识别算法（the license plate recognition algorithm based on neural network and morphology）》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、基于形态学和神经网络的车牌识别算法（Thelicenseplaterecognitionalgorithmbasedonneuralnetworkandmorphology）Twenty-fifthvolumethirdSeptember2001JournalofWuhanUniversityofTechnology(SCIENCEEDITION)JournalofWuhanUniversityofTechnology(TransportationScience&Engineering)Vol.25No.3September

2、2001ARecognitionofVehicleplatewithMathematicalMorphologyandNeuralNetworks1.YangJieLiXuGuoWei(CollegeofInformationTechnology,WUT,Wuhan430063)Abstract:RequirementofautomaticrecognitionofvehiclelicensenumbersfromtheirplatesisIncreasingforsocialservicessuchastrafficcon

3、trolandcollectionmotor-waycharges.ThispaperProposesanalgorithmbasedonmathematicalmorphologyoperationstodetecttheregionofaVehiclelicenceplatefromanimage.Theuniquenessofthisalgorithm,incomparisonwithotherApproachestothisproblem,isthattherearenospecialrestrictionsimpo

4、sedontheinputimages.Automaticrecognitionandclassificationbasedonfeedwardthreelayersneuralnetworksareused.Allalgorithmsinthepaperareverifiedatcomputerandexperimentalresultsare586efficientforRecognitionandclassification.Keywords:mathematicalmorphology;neuralnetworks;

5、recognition0IntroductionTrafficengineersandthepoliceoftenwishToabletouniquelyidentifyroadvehiclesatDifferentpointsinaroadnetworkforreasons,Suchas:1)automatictoll-collection,roadpricingSchemes,andoperationofcarparksforrevenueCollection;2)fleetcontrolformonitoringand

6、OptimizingthemovementsofparticulartypesofVehicles;3)speedandweight-limitenforcementPurposes;4detection/preventioncrime);5)trafficdatacollectionetc.Untilquiterecently,vehicleidentificationHaslargelybeenalabor-intensivemanualActivity,aidedwherepossiblebyElectromechan

7、icalactuators.WiththeContinuingincreaseinvolumeofroadtraffic,Suchmanualtechniqueshaveprovedimpractical.ImprovementsinelectronicsensortechnologyHaveallowedforthedevelopmentofautomaticVehicleidentification(AVI)equipment.ConventionalAVIsystemsemployMicrowaveandinducti

8、velooptechnologies.InSuchsystems,auniquecodeidentifyingeachVehicleisstoredonavehiclemountedTransponder.AsthevehiclepassesspecificPointsintheroadw