自然光照條件下基于機(jī)器視覺的番茄缺素的智能診斷研究.pdf

1、Y827874sf毛UDC注1：笪!E!堡垡蛾期擎JlANGSUUNJVERSlTY街級編號碩士學(xué)位論文MASTER’SDISSERTATION@業(yè)笛擎監(jiān)Q￡緲論文題目：且熊盤盟系緝玉鶴拉鍪隧盟芻趑盎曲盤絲重顰塑壘學(xué)科專業(yè)：～壘生蔓魚塑查l墮e匠雄作者姓名：!生!拯指導(dǎo)教師：—碑辱答辯日期：一!i：61塑～——江蘇大學(xué)碩士學(xué)位論文ABSTRACTBasedonreferringtomuchliteratme，theresearchoft

2、omatonutrientdeficiencyundernaturalilluminationwasputforwardTheimageswerecollectedbydigitalcameraintelligentlyThemainpurposeofthisstudyistOdiagnosetomatodiseaseintellectivelybyusingplantphysiologyestablishmentplanting，di

3、gitalimagesprocessingandpattemrecognitionThemaincontentsofthestudyareasfollows：(1)CultivatedpuritynutrientdeficiencyspecimenDeficiencyofN，Ga，MgandFeduringthegrowthtimeoftomatousedasexperimentalobjectsTheShanQimethodwasus

4、edforcultivatingpuritytomatonutrientdeficiency(2)Advancedanewmethod，supergreenaddedB，forsegmentingimagesfromcomplexseRing，whichiseffectivebyexperimentTheratioofmaximalsquaresubtractmethodWaSusedtocomputerthethreshold，whi

5、chwasnotbeendesignedtheparametersofman—madebutvariancedynamicmethod，Differentimagepreprocessingwayssuchasremovingnoiseandenhancingimageswereusedonimages(3)ForgeRingtheinputfeaturevectorsofpaRemrecognition，thecolorfeature

6、andtexturalfeaturesextractionwayswerediscussedinthestudyThepercentageofdiseaseleavesareatototalareaisacolorfeatureparameterThecolorfeatureincludesthepercentageofareaandthesquaresubtractmethodetcThetexturalfeatureofindivi

7、dualleaveswereextractedbydifferentmethodsSomewereextractedbygrads—graycooccurrenceandFourierismmethodSomewereextractedbythewaveletmethod，whichwasespeciallysuitableforthetimefrequencyregionfeatures(4)Arecognitionmodelford

8、eficiencyleaveswasbuiltbyusingfuzzyK—neighbormethodSelectedtheeffectivefeatureparametersaStheinputfeaturevectorsofpaaemrecognitionUsingtheabovecolorandtexturalfeatureasinputvectorsoftherecognitionsystem，theclassification

9、accuraciesofthetestingdatasetwere925％，75％，80％，70％and85％，respectively，fornutrientdeficiencyleaves，deficiencyofCa，deficiencyofMg，deficiencyofFe，deficiencyofNiTheresearchprovidesimportantbasictheoriesandtechniquesforfurther

10、investigationdevelopmentofplantdeficiencyrecognitionsystem，whichwasvaluableinbusinessapplianceItisofgreatsignificanceforpromotingtheapplicationofcomputevisioninagricultureengineeringfieldKeywords：Tomato；Nutritiondeficien