基于本體的材料腐蝕特征信息檢索模型及應(yīng)用技術(shù)研究.pdf

1、重慶理工大學(xué)碩士學(xué)位論文基于本體的材料腐蝕特征信息檢索模型及應(yīng)用技術(shù)研究姓名：張菲申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：紀(jì)鋼20100610AbstractIIAbstractNowadayscrosionacteristiconthesurfaceofmaterialismainlysharedadministerthroughthemethodoftextretrievalinaChinaIndustriesInstitu

2、teNaturalEnvironmentalTestingCenter.Astextretrievaloncrosionacteristicisinfluencedbyindividualdifferencesexperiencessotherearesomedefectswithit.Firstlyitishardtoguaranteethereliabilityofretrieval.Secondlythesimilarityofc

3、rosionconditionsmakesmanysimilarimagesintheimagelibrary.Finallyintheinfmationageastheincreasingnumberofcrosionfeatureimagesthekeywdretrievalmethodisalsodifficulttoadapttocustomerneeds.Ontologytheyhasbeendrawingmemeresear

4、chattentioninthefieldofimageretrievalinrecentyears.Inthisthesiswefirstlyintroduceontologytheythenbuiltageneralpurposeontologymodelusedtodescribeimagescomprehensivelywithwhichwecandescribedifferentaspectsofimageinfmationm

5、akefivetupledeionofontology.Itisdifficulttoundersttheimagedirectlyobtainedfromthelowlevelvisualfeatureswhichrequiresanalyzingthesemanticsoftheimagedeeplymakingthecomputercanunderstthesimilaritybasedonhumancognition.Tothi

6、sendweassociatethebodyproposedframewkfontologybasedimage:byontologybasedsemanticextensionretrievalsemanticqueryprocesstomakeupfthelackofinfmationthroughtheontologythesemanticdefinitionoftheconceptofimagerelationship.Imag

7、eontologymodelsincludethevisualimagefeatureshighlevelsemanticconceptnotonlytakefulladvantageofthelowlevelfeaturesofimageitselfbutalsoconsistentwiththevisualimagesofpeoplefromunderstingthusfillingthe“semanticgap“betweenth

8、elowlevelvisualfeatureshighlevelsemanticsotherstudiesstaffinspiredtodothispaperwithfurtherimprovementofwatershedalgithminensuringthesegmentationresultsbasicallyunchangedtolowerthetimecomplexity.Indertoachievehighlevelima

9、geautomaticextractionofsemanticfeaturesthispaperintroducesthebasictheyofSVMsupptvectmachineusingtheunderlyingacteristicsoftheimagemappedtotheontologyofthehighlevelsemanticconceptssowewillautomaticallyobtainsemanticannota