基于視覺的運(yùn)動目標(biāo)跟蹤系統(tǒng)研究.pdf

1、武漢科技大學(xué)碩士學(xué)位論文基于視覺的運(yùn)動目標(biāo)跟蹤系統(tǒng)研究姓名：郭靜申請學(xué)位級別：碩士專業(yè)：電路與系統(tǒng)指導(dǎo)教師：吳懷宇20120619AbstractMoVingobjectdetectionand虹acl(il唱isoneofmaillprojectsofcomputer、risionresearch，whichmsesmanyadvaulcedteclulologylike渤ageprocessillg，pattemrecognjtio

2、n，automaticcomol，artificialintelligenceandcomputerscienceandotherfieldsaIldithas誠despreadapplicationiIlmilit哪guidance，smansurveillaIlce，medicaldia印osis，仃a伍cmoIlito血gandrobotfield，ⅡmstheresearchoftllisprojecthaSgreatsigll

3、ificanceinnleaCadeIllicand印plicationValuehltbisdissertation，t11ekeyteclmologyofmisprojectlikeimagepreprocessing，objectdetectionandobject仃acHngwillbepresentedaIldthefollowiIlgist：hemaill、Ⅳork：Firstly，tllepresentstudystatu

4、sa11dmethodsindomesticandoverseaSofmovingobjectdetectionaIldtrackingares啪arizedandmeadvantageanddisadvalltageofeachme廿10darediscussed，w11ichestablishesⅡ1eimponaultresearchfIomldationfort11isprqectSecondly，mebasicprincipl

5、esofaveragefiltering，mediaIlfiltering，m砒lema：ticsmo印hoIogya工ldc011nectiV塒testingareintroduceda11dtheadVantageanddisadVantageofeachmetllodareanalyzedt11roughcontraStexperimentsTllirdly，鉚omemodsofimagesegmentationliketIle1

6、1istogr咖segmentationalgori也maIldmeOTSUmresholdsegmljIltationalgorithrnarepresentedThen也reemovingobjectdetectionmethodsl墩etheinter行鋤edifferencealgorimm，們hebackgromlddifferencingalgorimmandtheopticalnowmemodareelaborated11

7、1eme恤dofobjectdetectionusingbackgroundestiIIlationⅥ恤chisbaSedonGMMbackgrouIldupdatesalgorimmisacKeVedandt11eexpcrimemalresmtsshowmat也edetectedobjectiscompleteFinally，mepriIlciplesofMeanShiRalgorithmandKalmanfilterareelab

8、oratedThenthismesispmposesanewappmachcombinedMeallSKfttrackingalgod也mandKalInan矗lterwKchisappliedtoprcdictthemotioni11fomationofobjectisi11仃oducedandthe仃acI咄theobjectwllichhaschaIlgingsizeisacmeVedThenarlew印proachwllichi

9、sappliedt0仃ackmultipleobjectscombinedbIobmodelinga11dmodifiedMeanSmR仃ackillgalgoriⅡunispresentedandtheexperimentalresultsshowt11舭t11eapproachiseffeCtivea11drobustKqwords：ObjectDetection；ObjectTracking；ImageSegmentation；M