地下植物根莖探測(cè)技術(shù)研究.pdf

1、聲明本學(xué)位論文是我在導(dǎo)師的指導(dǎo)下取得的研究成果，盡我所知，在本學(xué)位論文中，除了加以標(biāo)注和致謝的部分外，不包含其他人已經(jīng)發(fā)表或公布過(guò)的研究成果，也不包含我為獲得任何教育機(jī)構(gòu)的學(xué)位或?qū)W歷而使用過(guò)的材料。與我一同工作的同事對(duì)本學(xué)位論文做出的貢獻(xiàn)均己在論文中作了明確的說(shuō)明。研究生簽名：月垢碩士學(xué)位論文壘!墅堅(jiān)———————————————————————————————一AbstractBasedonreq毗mentsofanenterpris

2、eproject，thedetectiontechnologyofu11de理印u11dDlantrootisstudiedinthispaperThemaintasksinclude：(1)nondestructivelytodetectobjedu】帕ergrouIld；(2)toidentifytheundergroundplantroot；(3)todeterminethes1笳ofplantmo‘Inthispapergro吼

3、dpenetratingradarisusedtonondestructivelydetectundergro衄dpl粗trootAccordiIlgtomeo№inedechodataofground—penetratingradarthem鋤pointmthlspaperisplacedonechosignalprocessingandundergroundplantrootidentificatlonFirstlv’accordi

4、ngtotherequirementsofplantrootdetection，thispaperaIlalyzeselec仃oma鱸eticwavepropagationcharacteristicsinsoilandthereso№utonofground供netratingradarandSOon；Atthesametime，electromagneticcharactensncsots01landplantrootalsoare

5、analyzedinordertosimulatethedetectionprocesso士madergrouIldobjectwithgroundpenetratingradarsimulationsoftwareSecondlv’thegroundpenetratingradarsignalprocessisstudiedThesecontentsmallllyi11cludetheGaussi狃noisereduction，und

6、ergroundobjectdetection，remoValofthedirectwaVeundergroundobjectsizeestimation，groundpenetratingradargroundecho鋤ValtimeestimationAftersignalprocessing，Ontheonehand，directwaveandn01seareSuppressed，nlerebvtheechoinfb肌ationo

7、fundergroundobjectenhancedsignificantly；Ontheotllerhalld，thedepthandsizeinformationofundergroundobjectareobtainedAgainundergroundsectionimageisreconstructedwiththetwo—dimensionalechodataofgroundpenetratingradarTwodiffere

8、ntmethodsofimagereconstructionareuSedmthispaperoneisbasedoncharaCteristicsofobjectshyperbolic，andanotherisbasedonstoltinterpolationAflerimagereconstruction，theazimuthresolutionofGPRhaSbeensignificantlyimproved一Finallvthe

9、identificationofundergroundplantrootisstudiedGroundpenetratingradarsimulationsoR咄isusedfirsttoobtaintheechodataofplantroot，stoneandotherobjects，thenfeaturevectorsareextractedfromtheseechoes，andsupportvectormacmneISadopte

10、dtoclass衙thesefeaturevectorsResultsshowthattherecognitionaccuracyrateofplantroot1sabove90％TherebytheidentificationmethoddescribedhereCanbeusedinthe代se盯chofidentificationofdifferentundergroundobjectsII碰M取叭黧礎(chǔ)‰眥卻啦一暑撕培m姍嘶∞試印