數(shù)字圖像技術(shù)在冬小麥群體監(jiān)測(cè)中的應(yīng)用研究.pdf

1、中國(guó)農(nóng)業(yè)大學(xué)博士學(xué)位論文數(shù)字圖像技術(shù)在冬小麥群體監(jiān)測(cè)中的應(yīng)用研究姓名：?jiǎn)纬射撋暾?qǐng)學(xué)位級(jí)別：博士專業(yè)：作物栽培學(xué)與耕作學(xué)指導(dǎo)教師：王璞廖樹華20060601中國(guó)農(nóng)業(yè)大學(xué)博士學(xué)位論文ABSTRACTABSTRACTThetheoryandtechniqueofdigitalimageprocessing(DIP)areappliedondevelopingthemethodsformonitoringwinterwheatcanopyund

2、erfieldconditionsThetechniquesofimageacquisition，imagesegmentationandcolorcalibrationaredevelopedFurthermore，theimagefeatureindicesofcanopyscale，canopynutritionalstatusandcanopycomponentsdistributionaredeterminedThemajor

3、resultsareasfollows：1Threestandardmethodsofimageacquisitionareestablished，whichareusedinmonitoringthescaleandtextureofcanopythenutritionalstatusofcanopycomponentsdistributionanduniformityofcanopy2Anewmethodofimagesegment

4、ationbasedonExGhistogramisimproved，andthecorrectclassificationrateincreasesfrom74％to88％Thehighlightareainwheatcanopyisabout08％TheprocessingmethodsarepresentedinordertoreduceitsinfluenceAftercolorbalanceaccordingtocolorof

5、greycard，varianceofhuevaluecouldbereducedfrom5655to4743Onthebasisoftheoryandexperimentaldata，thelogarithmicmodelbetweencanopycoverageandleafareaindexisestablished：LAl2—22441n(1一rc)，whichhastheexactbiologicalsignificancea

6、ndthedegreeoffittingisimprovedCanopy—coverage—duration(CcD)isusedfordescribingthecanopycoverageaccumulationduringaperiod；itstherioticalexpressionandcomputingfunctionaredevelopedEulernumberisintroducedonthestudyofwheatcan

7、opySignificantlynegativecorrelationbetweenEulernumberandLAIisobeservedbeforeheadingofwinterwheat；themodelsofregressionaredifferentbetweencultivars(Jin9411andJingDon98)atthesamegrowthstage，whilenosignificantdifferencefort

8、hesamecultivaramongtreatmentsTheimagemethodsbasedonimageedgepixelnumberaredevelopedtoestimatethetotalstemnumberofwheat，andthe3layersBPneuralnetworkisappliedinthemodel4Thetexturefeatureindexesaredeterminedbystepwisediscri

9、minantmethodThecorrectclassificationrateatbootingstageisthehighest(1OO％)comparingwiththatatothergrowthstagesTherefore，thedifferentcanopystructureofwinterwheatcouldbeclassifiedbyanalysisingimagetextureAnewimageindex(PUI)i