基于機(jī)器視覺(jué)的車輛距離測(cè)量技術(shù).pdf

1、南京理工大學(xué)碩士學(xué)位論文基于機(jī)器視覺(jué)的車輛距離測(cè)量技術(shù)姓名：王青申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：電子與通信工程指導(dǎo)教師：謝仁宏陸志全20081201AbstractRangingsystemisaIliInport觚tpanofIVS(Intelligentvehiclesystem)IthastheVitalsigIlific觚cet0travel’ssec耐tyeIlllaIlcementandtramcaccident’soccu玎enc

2、ereductionSonlcraJlgingmethodswildlyusedinpresentareaIlalyzedinme抓icleComparedmeirstrongpoints缸dSh硎ages，ar眥g婦喀methodbaSedonmaChineVisionisproposedItadoptsmonocularCCDvidicon嬲inputdeVice，aIldprocessestheinputimagebycomput

3、er，firlallycalculatesthedist|an∞accordingt0themodelondistancebetween鉚ovehicles1heentire啪gingsystemdesi驢isin仃0ducedinnlep印erincludingmesystemstrIlcture，theh油arcplatfoml，t：hesomⅣaresimulationplatfonn鶴well緦systcInIspe墑mance

4、，a11dtheral培ingalgo枷：lIIlprincipleandthenowiselaboratedwimemphaSisbaSedonmeimageprocessingtechnolog)rFirstlyt量lewavelett11eo巧isusedtocan了ontheimagedenoisingaIldtheimage翎lllancementn玲iterativethresholdSegmen協(xié)tionar砒lmetic

5、isproposedtosegmentthelalleandtlle壬@schoperatorisselectedtodetecttheimageedgeNeXt，meran百nga190ritllIIl’sincludingthreeimportalltstepsisrese2Lrchedde印lyRegardingmelallelirledetection，akindofimproVementmemodbaSedon如zzytheo

6、r)，isproposednoVerc伽∞stheshortcomingsofconVentionalHou曲衄lsformmethodThcexperinlentresultsshowmatthem甜lodisefI’ective觚ditcandeteCtt：helineaccuratelyThen，inordertofilldtllevellicleaccurateposition，themIetllodbasedonroads耐a

7、Ceaveragegmdation鋤dⅡ忙winderenergyarepre∞ntedFinally，晰thmeaid0fmegeometrictransf0徹t11eoⅨt11emonocular啪gingmodelisestablishedaIldthe1eadingVehicledistanceiscalculated1kexperimentindicatedtha土tIlisset0falgorithmisreal—timea