外文翻译图像的边缘检测.docx

外文翻译图像的边缘检测.docx

《外文翻译图像的边缘检测.docx》由会员分享，可在线阅读，更多相关《外文翻译图像的边缘检测.docx（5页珍藏版）》请在冰点文库上搜索。

附：

英文资料翻译

图像的边缘检测

Toimageedgeexaminationalgorithmresearchacademicreport

Abstract

Digitalimageprocessingtookarelativequiteyoungdiscipline,isfollowingthecomputertechnologyrapiddevelopment,daybydayobtainsthewidespreadapplication.Theedgetooktheimageonekindofbasiccharacteristic,inthepatternrecognition,theimagedivision,theimageintensificationaswellastheimagecompressionandsooninthedomainhasamorewidespreadapplication.Imageedgedetectionmethodmanyandvaried,inwhichbasedonbrightnessalgorithm,isstudiesthetimetobemostlong,thetheorydevelopsthematurestmethod,itmainlyisthroughsomedifferenceoperator,calculatesitsgradientbasedonimagebrightnessthechange,thusexaminestheedge,mainlyhasRobert,Laplacian,Sobel,Canny,operatorsandsoonLOG.Firstasawholeintroduceddigitalimageprocessingandtheedgedetectionsurvey,hasenumeratedseveralkindofatpresentcommonlyusededgedetectiontechnologyandthealgorithm,andselectstwokindstouseVisualtheClanguageprogrammingrealization,throughwithdrawstheimageresulttotwoalgorithmsthecomparison,theresearchdiscussestheirgoodandbadpoints.

对图像边缘检测算法的研究学术报告摘要

数字图像处理作为一门相对比较年轻的学科,伴随着计算机技术的飞速发展,日益得到广泛的应用.边缘作为图像的一种基本特征,在图像识别,图像分割,图像增强以及图像压缩等的领域中有较为广泛的应用.图像边缘提取的手段多种多样,其中基于亮度的算法,是研究时间最久,理论发展最成熟的方法,它主要是通过一些差分算子,由图像的亮度计算其梯度的变化,从而检测出边缘,主要有Robert,Laplacian,Sobel,Canny,LOG等算子.首先从总体上介绍了数字图像处理及边缘提取的概况,列举了几种目前常用的边缘提取技术和算法,并选取其中两种使用VisualC++语言编程实现,通过对两种算法所提取图像结果的比较,研究探讨它们的优缺点.

Firstchapterintroduction

§1.1imageedgeexaminationintroduction

Theimageedgeisoneofimagemostbasiccharacteristics,ofteniscarryingimagemajorityofinformations.Buttheedgeexistsintheimageirregularstructureandinnotthesteadyphenomenon,alsonamelyexistsinthesignalpointofdiscontinuityplace,thesespotshavegiventheimageoutlineposition,theseoutlinesarefrequentlywewhentheimageryprocessingneedstheextremelyimportantsomerepresentativecondition,thisneedsustoexamineandtowithdrawitsedgetoanimage.Buttheedgeexaminationalgorithmisintheimageryprocessingquestiononeofclassicaltechnicaldifficultproblems,itssolutioncarriesonthehighlevelregardingusthecharacteristicdescription,therecognitionandtheunderstandingandsoonhasthesignificantinfluence;Alsobecausetheedgeexaminationallhasinmanyaspectstheextremelyimportantusevalue,thereforehowthepeoplearedevotingcontinuouslyinstudyandsolvethestructuretoleavehavethegoodnatureandthegoodeffectedgeexaminationoperatorquestion.Intheusualsituation,wemaythesignalinsingularpointandthepointofdiscontinuitythoughtisintheimageperipheralpoint,itsnearbygradationchangesituationmayreflectfromitsneighboringpictureelementgradationdistributiongradient.Accordingtothischaracteristic,weproposedmanykindsofedgeexaminationoperator:

IfRobertoperator,Sobeloperator,Prewittoperator,Laplaceoperatorandsoon.Thesemethodsmanyarewaitfortheprocessingpictureelementtocarryonthegradationanalysisforthecentralneighborhoodachievementthefoundation,realizedandhasalreadyobtainedthegoodprocessingeffecttotheimageedgeextraction..Butthiskindofmethodsimultaneouslyalsoexistshastheedgepictureelementwidth,thenoisejammingisseriousandsoontheshortcomings,evenifusessomeauxiliarymethodstoperformthedenoising,alsocorrespondingcanbringtheflawwhichtheedgefuzzyandsoonovercomeswithdifficulty.Alongwiththewaveletanalysisappearance,itsgoodtimefrequencypartialcharacteristicbythewidespreadapplicationintheimageryprocessingandinthepatternrecognitiondomain,becomesinthesignalprocessingthecommonlyusedmethodandthepowerfultool.Throughthewaveletanalysis,mayinterweavedecomposesinthesameplaceeachkindofcompositesignalthedifferentfrequencytheblocksignal,butcarriesontheedgeexaminationthroughthewavelettransformation,mayuseitsmulti-criteriaandthemulti-resolutionnaturefully,realeffectiveexpressestheimagetheedgecharacteristic.Whenthewavelettransformationcriterionreduces,ismoresensitivetotheimagedetail;Butwhenthecriterionincreases,theimagedetailisfilteredout,theexaminationedgewillbeonlythethickoutline.Thischaracteristicisextremelyusefulinthepatternrecognition,wemaybecalledthisthickoutlinetheimagethemainedge.Ifwillbeableanimagemainedgeclearintegrityextraction,thistothegoaldivision,therecognitionandsoonfollowingprocessingtobringtheenormousconvenience.Generallyspeaking,theabovemethodallistheworkwhichdoesbasedontheimageluminanceinformation.Inthemultitudinousscientificresearchworkerunder,hasobtainedtheverygoodeffectdiligently.But,becausetheimageedgereceivesphysicalconditionandsoontheilluminationinfluencesquitetobebigabove,oftenenablesmanytohaveacommonshortcomingbasedonbrightnessedgedetectionmethod,thatistheedgeisnotcontinual,doesnotsealup.Consideredthephaseinformationintheimageimportanceaswellasitsstablecharacteristic,causesusingthephaseinformationtocarryontheimageryprocessingintonewresearchtopic.Inthispapersoonintroducesonekindbasedonthephaseimagecharacteristicexaminationmethod--phaseuniformmethod.Itisnotusestheimagetheluminanceinformation,butisitsphasecharacteristic,namelysuppositionimageFouriercomponentphasemostconsistentspotachievementcharacteristicpoint.Notonlyitcanexaminebrightnesscharacteristicsandsoonstepcharacteristic,linecharacteristic,moreovercanexamineMachbeltphenomenonwhichproducesasaresultofthehumanvisionsensationcharacteristic.Becausethephaseuniformitydoesnotneedtocarryonanysuppositiontotheimagecharacteristictype,thereforeithastheverystrongversatility.

第一章绪论

§1.1图像边缘检测概论

图像边缘是图像最基本的特征之一,往往携带着一幅图像的大部分信息.而边缘存在于图像的不规则结构和不平稳现象中,也即存在于信号的突变点处,这些点给出了图像轮廓的位置,这些轮廓常常是我们在图像处理时所需要的非常重要的一些特征条件,这就需要我们对一幅图像检测并提取出它的边缘.而边缘检测算法则是图像处理问题中经典技术难题之一,它的解决对于我们进行高层次的特征描述,识别和理解等有着重大的影响;又由于边缘检测在许多方面都有着非常重要的使用价值,所以人们一直在致力于研究和解决如何构造出具有良好性质及好的效果的边缘检测算子的问题.在通常情况下,我们可以将信号中的奇异点和突变点认为是图像中的边缘点,其附近灰度的变化情况可从它相邻像素灰度分布的梯度来反映.

根据这一特点,我们提出了多种边缘检测算子:

如Robert算子,Sobel算子,Prewitt算子,Laplace算子等.这些方法多是以待处理像素为中心的邻域作为进行灰度分析的基础,实现对图像边缘的提取并已经取得了较好的处理效果.但这类方法同时也存在有边缘像素宽,噪声干扰较严重等缺点,即使采用一些辅助的方法加以去噪,也相应的会带来边缘模糊等难以克服的缺陷.随着小波分析的出现,其良好的时频局部特性被广泛的应用在图像处理和模式识别领域中,成为信号处理中常用的手段和有力的工具.通过小波分析,可以将交织在一起的各种混合信号分解成不同频率的块信号,而通过小波变换进行边缘检测,可以充分利用其多尺度和多分辨率的性质,真实有效的表达图像的边缘特征.当小波变换的尺度减小时,对图像的细节更加敏感;而当尺度增大时,图像的细节将被滤掉,检测的边缘只是粗轮廓.该特性在模式识别中非常有用,我们可以将此粗轮廓称为图像的主要边缘.如果能将一个图像的主要边缘清晰完整的提取出来,这将对目标分割,识别等后续处理带来极大的便利.总的说来,以上方法都是基于图像的亮度信息来作的工作.在众多科研工作者的努力下,取得了很好的效果.但是,由于图像边缘受到光照等物理条件的影响比较大,往往使得以上诸多基于亮度的边缘提取方法有着一个共同的缺点,那就是边缘不连续,不封闭.考虑到相位信息在图像中的重要性以及其稳定的特点,使得利用相位信息进行图像处理成为新的研究课题.在本文中即将介绍一种基于相位的图像特征检测方法——相位一致性方法.它并不是利用图像的亮度信息,而是其相位特点,即假设图像的傅立叶分量相位最一致的点作为特征点.它不但能检测到阶跃特征,线特征等亮度特征,而且能够检测到由于人类视觉感知特性而产生的的马赫带现象.由于相位一致性不需要对图像的特征类型进行任何假设,所以它具有很强的通用性.

§1.2imageedgedefinition

Theimagemajoritymaininformationallexistsintheimageedge,themainperformancefortheimagepartialcharacteristicdiscontinuity,isintheimagethegradationchangequitefierceplace,alsoisthesignalwhichweusuallysaidhasthestrangechangeplace.Thestrangesignalthegradationchangewhichmovestowardsalongtheedgeisfierce,usuallywedividetheedgeforthestepshapeandtheroofshapetwokindoftypes（asshowninFigure1-1）.Inthestepedgetwosidegreylevelshavetheobviouschange;Buttheroofshapeedgeislocatedthegradationincreaseandthereducedintersectionpoint.Mayportraytheperipheralpointinmathematicsusingthegradationderivativethechange,tothestepedge,theroofshapeedgeasksitsstep,thesecondtimederivativeseparately.Toanedge,hasthepossibilitysimultaneouslytohavethestepandthelineedgecharacteristic.Forexampleonasurface,changesfromaplanetothenormaldirectiondifferentanotherplanecanproducethestepedge;Ifthissurfacehastheedgesandcornerswhichtheregularreflectioncharacteristicalsotwoplanesformquitetobesmooth,thenworksaswhenedgesandcornerssmoothsurfacenormalaftermirrorsurfacereflectionangle,asaresultoftheregularreflectioncomponent,canproducethebrightlightstripontheedgesandcornerssmoothsurface,suchedgelookedlikehaslikelysuperimposedalineedgeinthestepedge.Becauseedgepossibleandinsceneobjectimportantcharacteristiccorrespondence,thereforeitistheveryimportantimagecharacteristic.Forinstance,anobjectoutlineusuallyproducesthestepedge,becausetheobjectimageintensityisdifferentwiththebackgroundimageintensity.

§1.3paperselectedtopictheorysignificance

Thepaperselectedtopicoriginatesinholdstheimportantstatusandthefunctionpracticalapplicationtopicintheimageproject.Theso-calledimageprojectdisciplineisrefersfoundationdisciplineandsoonmathematics,opticsprinciples,thedisciplinewhichintheimageapplicationunifieswhichaccumulatesthetechnicalbackgrounddevelops.Theimageprojectcontentisextremelyrich,andsoondividesintothreelevelsdifferentlyaccordingtotheabstractdegreeandtheresearchtechnique:

Imageryprocessing,imageanalysisandimageunderstanding.AsshowninFigure1-2,inthecharttheimagedivisionisinbetweentheimageanalysisandtheimageryprocessing,itsmeaningis,theimagedivisionisfromtheimageryprocessingtotheimageanalysisessentialstep,alsoisfurtherunderstandstheimagethefoundation.Theimagedivisio