数字图像处理外文翻译参考文献Word文档下载推荐.docx

数字图像处理外文翻译参考文献Word文档下载推荐.docx

《数字图像处理外文翻译参考文献Word文档下载推荐.docx》由会员分享，可在线阅读，更多相关《数字图像处理外文翻译参考文献Word文档下载推荐.docx（13页珍藏版）》请在冰豆网上搜索。

原文：

ApplicationOfDigitalImageProcessingInTheMeasurementOfCastingSurfaceRoughness

Ahstract-Thispaperpresentsasurfaceimageacquisitionsystembasedondigitalimageprocessingtechnology.TheimageacquiredbyCCDispre-processedthroughtheprocedureofimageediting,imageequalization,theimagebinaryconversationandfeatureparametersextractiontoachievecastingsurfaceroughnessmeasurement.Thethree-dimensionalevaluationmethodistakentoobtaintheevaluationparameters

andthecastingsurfaceroughnessbasedonfeatureparametersextraction.AnautomaticdetectioninterfaceofcastingsurfaceroughnessbasedonMATLABiscompiledwhichcanprovideasolidfoundationfortheonlineandfastdetectionofcastingsurfaceroughnessbasedonimageprocessingtechnology.

Keywords-castingsurface;

roughnessmeasurement;

imageprocessing;

featureparameters

Ⅰ.INTRODUCTION

Nowadaysthedemandforthequalityandsurfaceroughnessofmachiningishighlyincreased,andthemachinevisioninspectionbasedonimageprocessinghasbecomeoneofthehotspotofmeasuringtechnologyinmechanicalindustryduetotheiradvantagessuchasnon-contact,fastspeed,suitableprecision,strongabilityofanti-interference,etc[1,2].Asthereisnolawsaboutthecastingsurfaceandtherangeofroughnessiswide,detectionparametersjustrelatedtohighlydirectioncannotmeetthecurrentrequirementsofthedevelopmentofthephotoelectrictechnology,horizontalspacingorroughnessalsorequiresaquantitativerepresentation.Therefore,thethree-dimensionalevaluationsystemofthecastingsurfaceroughnessisestablishedasthegoal[3,4],surfaceroughnessmeasurementbasedonimageprocessingtechnologyispresented.Imagepreprocessingisdeducedthroughtheimageenhancementprocessing,theimagebinaryconversation.Thethree-dimensionalroughnessevaluationbasedonthefeatureparametersisperformed.AnautomaticdetectioninterfaceofcastingsurfaceroughnessbasedonMATLABiscompiledwhichprovidesasolidfoundationfortheonlineandfastdetectionofcastingsurfaceroughness.

II.CASTINGSURFACEIMAGEACQUISITIONSYSTEM

Theacquisitionsystemiscomposedofthesamplecarrier,microscope,CCDcamera,imageacquisitioncardandthecomputer.Samplecarrierisusedtoplacetestedcastings.Accordingtotheexperimentalrequirements,wecanselectafixedcarrierandthesamplelocationcanbemanuallytransformed,orselectcuringspecimensandthepositionofthesamplingstagecanbechanged.Figure1showsthewholeprocessingprocedure.,Firstly,thedetectedcastingsshouldbeplacedintheilluminatedbackgroundsasfaraspossible,andthenthroughregulatingopticallens,settingtheCCDcameraresolutionandexposuretime,thepicturescollectedbyCCDaresavedtocomputermemorythroughtheacquisitioncard.Theimagepreprocessingandfeaturevalueextractiononcastingsurfacebasedoncorrespondingsoftwarearefollowed.Finallythedetectingresultisoutput.

III.CASTINGSURFACEIMAGEPROCESSING

Castingsurfaceimageprocessingincludesimageediting,equalizationprocessing,imageenhancementandtheimagebinaryconversation,etc.TheoriginalandclippedimagesofthemeasuredcastingisgiveninFigure2.Inwhicha）presentstheoriginalimageandb）showstheclippedimage.

A.ImageEnhancement

Imageenhancementisakindofprocessingmethodwhichcanhighlightcertainimageinformationaccordingtosomespecificneedsandweakenorremovesomeunwantedinformationsatthesametime[5].Inordertoobtainmoreclearlycontourofthecastingsurfaceequalizationprocessingoftheimagenamelythecorrectionoftheimagehistogramshouldbepre-processedbeforeimagesegmentationprocessing.Figure3showstheoriginalgrayscaleimageandequalizationprocessingimageandtheirhistograms.Asshowninthefigure,eachgraylevelofthehistogramhassubstantiallythesamepixelpointandbecomesmoreflataftergrayequalizationprocessing.Theimageappearsmoreclearlyafterthecorrectionandthecontrastoftheimageisenhanced.

Fig.2Castingsurfaceimage

Fig.3Equalizationprocessingimage

B.ImageSegmentation

Imagesegmentationistheprocessofpixelclassificationinessence.Itisaveryimportanttechnologybythresholdclassification.Theoptimalthresholdisattainedthroughtheinstmctionthresh=graythresh（II）.Figure4showstheimageofthebinaryconversation.ThegrayvalueoftheblackareasoftheImagedisplaystheportionofthecontourlessthanthethreshold（0.43137）,whilethewhiteareashowsthegrayvaluegreaterthanthethreshold.Theshadowsandshadingemergeinthebrightregionmaybecausedbynoiseorsurfacedepression.

Fig4Binaryconversation

IV.ROUGHNESSPARAMETEREXTRACTION

Inordertodetectthesurfaceroug