基于模糊理論的倉(cāng)儲(chǔ)物昆蟲的分類識(shí)別研究.pdf

1、鄭州大學(xué)碩士學(xué)位論文基于模糊理論的倉(cāng)儲(chǔ)物昆蟲的分類識(shí)別研究姓名：劉素華申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)軟件與理論指導(dǎo)教師：陳國(guó)勛20031116幕于模糊理論的倉(cāng)儲(chǔ)物昆蟲的分類識(shí)別研究ABSTRATstoredDroductedinsects。arethemainobjectsintheeconomicalinsectsresearchThetargetwhatitencroachesoninvolvesvariouskindsofgood

2、sandmaterialsthathumanlifeneedsWithourrapideconomicdevelopment，theinternationaltradeisbecomingmorefrequentwhichobjectivelyofferadvantageousconditionforthesestoredproductedinsectstopropagateandspreadSostrengtheningthequar

3、antineofstoredproductedinsectshasimportantsignificancetaisethetechnologicallevelofpreventingandkillingthem，toprotectstoredmaterial，tolessonlossandtospeedupthenationaleconomicdevelopmentFormanyyears，theinsect’sresearchers

4、havebeenmakinggreateffortstolookforakindofscientific，fast，accurate，practicalmethodforappraisingthekindofstoredproductedinsectsInrecentyears，alongwiththeexpandingofcomputertechnologyapplicatedfield，thetechnologyofimagepro

5、cessingandthemethodsofpatternrecognitionmadeabreakthroughAtpresent，utilizingcomputertechnologytodiscernandclassifystoredproductedinsectsautomaticallyisoneoftheimportantsubjectininsects’researchfieldThispapertriestoclassi

6、fyanddiscernthestoredproductedinsectswiththefuzzysettheoryThefuzzysettheoryisputforwardbytheautocontrolexpertofUSALAZadehprofessorin1965Althoughitisraisedrelativelylate，itsapplicationsineachfieldareveryactivelyatpresentT

7、hispaperintroducesbrieflytheprincipleofthefuzzysettheoryandexplainsonelearningmethodwithoutsupervisedindetail，andtwokindsofthefuzzypatterrecognitionmethodsFirstthispaperdoessomepreprocessingwithenhancingimagebymedianfilt

8、eringandthehistogrambalancingtechnologyThenitusescomputertoextractandunify17characteristicsincludingthegreystatisticalfeature，thetexturefeatureandformfeaturefromthepreprocessedstoredproductedinsects’imageautomaticallyThe

9、nitformsthestandardpatterstorehousebytheimprovedfuzzycmeansalgorithm，anddesignsclassificationmachinebyfuzzypatterrecognitionODthebasisofapproximationvalueandtheprincipleofselectingthenearAtlastitputsthesubsequencestoredp

10、roductedinsectsintotheclassificationmachineandidentifiesthemWiththehelpofpestresearchgroupinZhengZhouinstituteoftechnologyIobtainninepiecesofspecimenimagethreekindsofstoredDroductedinsects’inthecivilasthestudysampleofcla

11、ssificationmachineandthenutilizemicro—cameraofopticsgather30piecesofstoredproductedimage，Usingthemethodsmentionedabovetodiscern，thediscerningrateisalwaysupto88％Ifeelquitesatisfiedwiththeanalyticresultsofinitialexperiment

12、aldataTheresearchofthisprojecthaswideapplicationprospectandpresentanewapproachtoidentifyandclassifystoredproductedinseetsKeyWords：StoredproductedInsects，ImageEnhancement，F(xiàn)eatureSelectionThePatternRecognition，F(xiàn)uzzySet，F(xiàn)uz