遙感技術(shù)在數(shù)字流域中的應(yīng)用研究.pdf

1、華中科技大學(xué)碩士學(xué)位論文遙感技術(shù)在數(shù)字流域中的應(yīng)用研究姓名：伍振軍申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：水利水電工程指導(dǎo)教師：劉廣潤劉吉平2002.5.10華中科技大學(xué)碩士學(xué)位論文AbstractTheaimofthispaperistOworkinthestartupoftheprojectofthedigitalQingjiang，probeintothefields，approaches，effect，cBi“ren$situationandp

2、otentialofremotesensingtechniquesinapplyingthedigitalvalleyevaluatethenecessityfeasibilityofremotesensingtechniquesinobtainingrapidlythebasalgeographicalinformationandsomeparametersoftheearth’ssurfaceinthedi畫udvalleyThed

3、etectionoflandcover|landusepresentsituationsisanalyzedfromseveralfieldssuchasthestatusoflandCover／landuseathomeandabroad，problemsexistingintheapplicationofthedynamicdetectionoflandusechange，currentsituationstudyoflanduse

4、inourcountryandseveralrepresentativeclassificationmethodsoflandcover／landuseSeveralclassificationmethodsforIandcovefflandusebaseonremotesensingtechnologyhaveintroducedandcontrastedindetailFromtheexperimentalresultsitCanb

5、eseenthattheclassificationmethodoflandcover／landUSebasedonlterativeSelfOrganizingDataAnalysisTechniqueAlgorithm(ISODATA)，inwhichtheKLtransformisusedtocarryoutfeatureextractionforaremotesensingimage，improvestheclassificat

6、ionaccuracyandsavestimespendingconsiderablyThispaperputsforwardanewclassificationmethodforlandcover／landuseinthedigitalvalleybasedonthecombinationsofgeneticalgorithmsandabackpropagationalgorithmThismethodmakesnseoftwolay

7、erKohonen’Sselforganizingfeaturemapnetworktoselectroughlytrainingsamplesforalaterthree—layerbackpropagationnetworkSomesamplepixelsthataleclassifiedincorrectlyareeliminatedbymealsofassistantinformationsuchaslandnsemaps，ge