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，更多相关内容在工程资料-天天文库。
1、外文资料EdgeFeatureExtractionBasedonDigitalImageProcessingTechniquesI.INTRODUCTIONTheedgeisasetofthosepixelswhosegreyhavestepchangeandrooftopchange,anditexistsbetweenobjectandbackground,objectandobject,regionandregion,andbetweenclementandclement.Edgealwaysindwellsin
2、twoneighboringareashavingdifferentgreylevel.Itistheresultofgreylevelbeingdiscontinuous.Edgedetectionisakindofmethodofimagesegmentationbasedonrangenon-continuity.Imageedgedetectionisoneofthebasalcontentsintheimageprocessingandanalysis,andalsoisakindofissueswhicha
3、reunabletoberesolvedcompletelysofar.Whenimageisacquired,thefactorssuchastheprojection,mix,aberranceandnoiseareproduced.Thesefactorsbringonimagefeature'sbluranddistortion,consequentlyitisverydifficulttoextractimagefeature.Moreover,duetosuchfactorsitisalsodifficul
4、ttodetectedge.Themethodofimageedgeandoutlinecharacteristicsdetectionandextractionhasbeenresearchhotinthedomainofimageprocessingandanalysistechnique.Edgefeatureextractionhasbeenappliedinmanyareaswidely.Thispapermainlydiscussesaboutadvantagesanddisadvantagesofseve
5、raledgedetectionoperatorsappliedinthecableinsulationparametermeasurement.Inordertogainmorelegibleimageoutline,firstlytheacquiredimageisfilteredanddenoised.Intheprocessofdenoising,wavelettransformationisused.Andthendifferentoperatorsareappliedtodetectedgeincludin
6、gDifferentialoperator,Logoperator,CannyoperatorandBinarymorphologyoperator.Finallytheedgepixelsofimageareconnectedusingthemethodofborderingclosed.Thenaclearandcompleteimageoutlinewillbeobtained.II.IMAGEDENOISINGAsweallknow,theactualgatheredimagescontainnoisesint
7、heprocessofformation,transmission,receptionandprocessing.Noisesdeterioratethequalityoftheimage.TheymakeimageblueAndmanyimportantfeaturesarecoveredup.Thisbringslotsofdifficultiestotheanalysis.Therefore,themainpurposeistoremovenoisesoftheimageinthestageofpretreatm
8、ent.Thetraditionaldenoisingmethodistheuseofalow-passorband-passfiltertodenoise.Itsshortcomingisthatthesignalisblurredwhennoisesareremoved.Thereisirreconcilablecontrad
