资源描述：
《基于数字图像处理技术的边缘特征提取-毕业论文外文翻译》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、外文资料EdgeFeatureExtractionBasedonDigitalImageProcessingTechniquesI.INTRODUCTIONTheedgeisasetofthosepixelswhosegreyhavestepchangeandrooftopchange,anditexistsbetweenobjectandbackground,objectandobject,regionandregion,andbetweenclementandclement.Edgealwaysindwells
2、intwoneighboringareashavingdifferentgreylevel.Itistheresultofgreylevelbeingdiscontinuous.Edgedetectionisakindofmethodofimagesegmentationbasedonrangenon-continuity.Imageedgedetectionisoneofthebasalcontentsintheimageprocessingandanalysis,andalsoisakindofissueswh
3、ichareunabletoberesolvedcompletelysofar.Whenimageisacquired,thefactorssuchastheprojection,mix,aberranceandnoiseareproduced.Thesefactorsbringonimagefeature'sbluranddistortion,consequentlyitisverydifficulttoextractimagefeature.Moreover,duetosuchfactorsitisalsodi
4、fficulttodetectedge.Themethodofimageedgeandoutlinecharacteristic'sdetectionandextractionhasbeenresearchhotinthedomainofimageprocessingandanalysistechnique.Edgefeatureextractionhasbeenappliedinmanyareaswidely.Thispapermainlydiscussesaboutadvantagesanddisadvanta
5、gesofseveraledgedetectionoperatorsappliedinthecableinsulationparametermeasurement.Inordertogainmorelegibleimageoutline,firstlytheacquiredimageisfilteredanddenoised.Intheprocessofdenoising,wavelettransformationisused.Andthendifferentoperatorsareappliedtodetecte
6、dgeincludingDifferentialoperator,Logoperator,CannyoperatorandBinarymorphologyoperator.Finallytheedgepixelsofimageareconnectedusingthemethodofborderingclosed.Thenaclearandcompleteimageoutlinewillbeobtained.II.IMAGEDENOISINGAsweallknow,theactualgatheredimagescon
7、tainnoisesintheprocessofformation,transmission,receptionandprocessing.Noisesdeterioratethequalityoftheimage.Theymakeimageblur.Andmanyimportantfeaturesarecoveredup.Thisbringslotsofdifficultiestotheanalysis.Therefore,themainpurposeistoremovenoisesoftheimageinthe
8、stageofpretreatment.Thetraditionaldenoisingmethodistheuseofalow-passorband-passfiltertodenoise.Itsshortcomingisthatthesignalisblurredwhennoisesareremoved.Thereisirreconcilablecontr
