運(yùn)動(dòng)人體的跟蹤及頭部定位系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn).pdf

1、電子科技大學(xué)碩士學(xué)位論文運(yùn)動(dòng)人體的跟蹤及頭部定位系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn)姓名：楊春麗申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：魯曉軍20090501ABSTRACT一——————————————————————————一ABSTRACT弧mme岫dopm饑tofmodemsocietyandpeople’sincreasingSellseofs刪慨hl觸匝id鋤ti夠rec0鰣tionisindispensableinmoreandmor

2、epubhcpl∞鷦跚蛐髂豫ilwaystalioIls，卻orts，banks，administrations，communitiesand∞蛆B10m咖岱舢鰣怕瑪anovelkindofidentityrecognitiontechnologies，Canid∞缸黟hum啦刪tyaCcomlgt0thephysiologicalorbehavioralcharactersandisbetter也觚似瞄。誠rec0鰣tionmetho

3、dsinaccuracyandspeedBiometricfe批su咖currentlyincludefingerprintfaceandiris，etcIllnlisdiss刪鴨a刪ectonfacerecognitionofnational863Pr01jcct螄a蜊州hod鋤dte澌quc0nthebasisofVisualPerceptionandCognitiOnmech眥鋤Mul_timodeQlaracterFe蜘auth

4、enticationwillbediscussedAspreⅥouSwork眥也is劬ject，hlⅡnaIlmotiontrackingandheadlocatingbased011videowill幽瞅IyaffecttheperformanceoffacedetectionThetopicsaremainlydiscussedinthisdissertationasfollows：detectionofmoⅥngtarget艦c‰

5、gofmo咖g0bj嘲趾dhead10洳辱，obectdet枷onwhichseparatesthemovingobjectfromthesequ鉍ti1Ilages1sthefo伽捌onofsubsequenttreatmentsBasingontheanalysisofthethree伽咖nlyal眢。珊吼ofmov鋤enttargetextraction(backgrounddifference，interfr鋤edin嗍ce趾d

6、opticalflowmethod)atthisstage，andtakingtheapplicationoccasionintoa洶吼t，backgrounddi觸cemethodisusedIntheobjectregionobtainedbYthemethoddiscussedabove，somegapsstillexist，andnoisesarealsofoundinthebackgrollndWithmeCombiningo

7、fmathematicalmorphologyandconnectivitydetectlon【scalefilter)，wecanreducethenoiseandsmoothoutgapsDuringthephaseofmovingtargetclassification，thedispersednessandareainR舢a(chǎn)№nofthemovingtargetisusedtecognizehumaninthisdiss眥atl

8、on0nmeresearchofobjecttrackingbystudyingseveralrelatedalgorithm，thisdissenationimprovestheextendKalmanfiltertotrackingtargetsIt唧嬲c眥roldposition肌dareateplacecomplicatedtemplatematching，thusthismethod仰proVesthetrackingspee