運(yùn)動(dòng)目標(biāo)的圖像識(shí)別與跟蹤研究.pdf

1、分類號(hào)互型Z!：IUDCb凸!堡j)碩士學(xué)位論文密級(jí)——編號(hào)衛(wèi)學(xué)少運(yùn)動(dòng)目標(biāo)的圖像識(shí)別與跟蹤研究TheResearchofImageRecognitionandTrackingofMoving0bject指導(dǎo)教師塑衛(wèi)星副數(shù)授作者姓名揚(yáng)耿申請(qǐng)學(xué)位級(jí)別巫士學(xué)科(專業(yè))控劍理論皇控劍工程論文提交日期2QQ生壘且論文答辯日期2QQ5生魚自學(xué)位授予單位和日期江蒸太堂2QQ5生旦答辯委員會(huì)主席垂!：皇望評(píng)閱人江蘇大學(xué)碩士學(xué)位論文AbstractThe

2、dynamicimageofmovingobjectoffersmoreinformationthanwhatsingleimagecanprovideThroughanalysisofmanyflamesofimage，weCallgaininformationwhichwecannotgetfromsingleimageImagerecognitionandtrackingofmovingobjectisaprocessofdete

3、ctingrecognition—tracking，whichisbasedondynamicimageanalysisandcombinemethodsofimagerecognitionandimagetrackingThisarticlechoosesthemovingobjectasthetargetItputsforwardamethodwhichcarryoutimagerecognitionandtrackingonmov

4、ingobjectinthesimplebackgroundandmakescomparisonandstudyonalgorithmofobjectrecognitionandtrackingItmakesprogressonalgorithmofimagepreprocessing，imagesegmentation，featureextracting，imagerecognition，trackingandpredicting，w

5、hichisbasedonanalysisofsystemerrorDuringimagerecognition，itimprovesthetraditionalfilteralgorithmusingcorrelativeinformationbetweenadjacentwindows，andenhancesrealtimepropertyWhat’smore，thisarticlecombinesOtsuandselfadapti

6、vethresholdsegmentationandapplyittoimagesegmentation，thengetsbettereffectThisarticlealsomakessimulationandprogrammingonkindsofalgorithms，andoptimizethelastchosenalgorithm，whichmaketheprogrammingcompactandreliable，itenhan

7、cesthestabilityandthereal—timepropertygreatlyAtthesametime，itmakesastudyonwindowadaptingalgorithmwhichisusedtotrackmovingobjectandenhancesthetrackingprecisionfurthermoreTheexperimentalresultsindicatethealgorithmsoperateq

8、uicklyandmeetthecommandofrealtimeproperty，andsimpletoberealizedthrow【ghhardwareWhat’Smore，whenobjectisshelteredforabrieftime，itstillCanmakepredictingandtrackingonmovingobject，anditachievestheprospectiveeffectKeyword：imag