简介：WepresentaninitialstudyoftheobjectfeaturesofOracle9i-thefirstofthemarket-leadingobject-relationaldatabasesystemsthatsupportsatrueobjectmodelontheserversideaswellasanODMG-styleC++languagebindingontheclientside.WediscusshowthesefeaturescanbeusedtoprovidepersistentobjectstorageintheHEPenvironment.

简介：

简介：<正>Manypeoplehurttheirbackswhentheytrytoliftheavythingsfromthefloor.Itiseasytohurtyourbackmuscleswhenyoupickupaheavyobject.However,thereisacorrectwaytoliftthingsfromthefloor.Ifyoupickupbigorheavyobjectscorrectly,youprobablywillnothurtyourback.

简介：TheCMSOOreconstructionprogram-ORCA-hasbeenusedsince1999toproducelargesamplesofreconstructedMonte-Carloeventsfordetectoroptimization,triggerandphysicsstudies,TheeventsarestoredinseveralObjectivityfederationsatCERN,intheUS,Italyandothercountries.ToperformtheirstudiesphysicistsusedifferenteventsamplesrangingfromcompletedatasetsofTBytesizetoonlyafeweventsoutofthesedatasets.WedescribetheimplementationoftheserequirementsintheORCAsoftwareandthewaycollctionsofeventsareaccessedforreading,writingorcopying.

简介：TheBaBerexperimentstoresitsdatainanObjectorientedfederateddatabasesuppliedbyObjectivity/DB(tm),thisdatabaseissurrently350TBinsizeandisexpectedtoincreaseconsiderablyastheexperimentmatures.Managementofthisdatabaserequirescarefulplanningandspecializedtoolsinordertomakethedataavailabletophysicistsinanefficientandtimelymanner,WediscusstheoperationalissuesandmanagementtoolsthatweredevelopedduringthepreviousruntodealwiththisvastquantityofdataatSLAC.

简介：部件相关性是一项重要软件措施。它是直接与软件可理解性，可维护性，和可重用性有关。在描述部件相关性的二个重要参数是在二个部件之间联合的类型和依赖部件的类型。取决于联合和依赖部件的类型的不同类型，能部件维护和部件复用上有不同效果。在这篇论文，我们把依赖部件划分成三种类型。我们然后分类各种各样的部件相关性并且在维护和复用上分析他们的效果。基于分类，我们在场一个相关性度量标准并且在11个公开源码的Java部件上验证它。我们的学习证明强壮的关联在再使用部件的部件和努力的相关性的测量之间存在。这显示部件相关性和建议度量标准的分类能进一步被用来代表另外的外部软件质量因素。电子增补材料这篇文章(doi:10.1007/s11390-007-9058-y)的联机版本contatins增补材料，它对授权用户可得到。

简介：Ghostimagingcouldbeusedtomakeaquickidentificationoforthogonalobjectsbymeansofphotocurrentcorrelationmeasurements.Inthispaper,weextendthemethodtoidentifynonorthogonalobjects.Inthemethod,anobjectisilluminatedbyonephotonfromanentangledpair,andtheotheroneisdiffractedintoaparticulardirectionbyapre-establishedmultiple-exposurehologramintheidlerarm.Bythecorrelationmeasurements,thenonorthogonalobjectinthesignalarmcouldbediscriminatedwithinaveryshorttime.Theconstraintsfortheidentificationofnonorthogonalobjectsarepresented,whichshowthatthenonorthogonalobjectscanbediscriminatedwhentheoverlappingportionbetweenanytwoobjectsislessthanhalfofalltheobjectsintheset.Thenumericalsimulationsfurtherverifytheresult.

简介：Aviewinobjectorienteddatabasescorrespondstovirtualschemawithrestructuredgeneralizationanddecompositionhierarchies,Numbersofviewcreationmethodologieshavebeenproposed.Amajordrawbackofexistingmethodologiesisthattheydonotmaintaintheclosureproperty.Thatis,theresultofaquerydoesnothavethesamesemanticsasembodiedintheobjectorienteddatamodel.Teherefore,thispaperpresentsaviewcreationmethodologythatderivesaclassinresponsetoauser'squery,integratesderivedclassinglobalschema(i.e.,considerstheproblemofclassesmovinginclasshierarchy)andselectstherequiredclassesfromglobalschematocreatetheviewforuser'squery.Novelideaofviewcreationincludes:(a)anobjectalgebraforclassderivationandcustomization(wherethederivedclassesintermsofobjectinstancesandprocedure/methodsarestudied),(b)maintenanceofclosureproperty,and(c)classificationalgorithmwhichprovidesmechanismtodealwiththeproblemofaclassmovinginaclasshierarchy.

简介：Compressivesensingisarevolutionaryideaproposedrecentlytoachievemuchlowersamplingrateforsignals.Intheimageapplicationwithlimitedresourcesthecameradatacanbestoredandprocessedincompressedform.Analgorithmformovingobjectandregiondetectioninvideousingacompressivesamplingisdeveloped.Thealgorithmestimatesmotioninformationofthemovingobjectandregionsinthevideofromthecompressivemeasurementsofthecurrentimageandbackgroundscene.Thealgorithmdoesnotperforminversecompressiveoperationtoobtaintheactualpixelsofthecurrentimagenortheestimatedbackground.Thisleadstoacomputationallyefficientmethodandasystemcomparedwiththeexistingmotionestimationmethods.Theexperimentalresultsshowthatthesamplingratecanreduceto25%withoutsacrificingperformance.

简介：Thispaperproposesanewmethodbasedonprincipalcomponentanalysistofindthedirectionofanobjectinanypose.Experimentsshowthatthismethodisfast,canbeappliedtoobjectswithanypixeldistributionandkeeptheoriginalpropertiesofobjectsinvariant.ItisanewapplicationofPCAinimageanalysis.

简介：Thispaperintroducesanoveltechniqueforobjectdetectionusinggeneticalgorithmsandmorphologicalprocessing.Themethodemploysakindofobjectorientedstructureelement,whichisderivedbygeneticalgorithms.Thepopulationofmorphologicalfiltersisiterativelyevaluatedaccordingtoastatisticalperformanceindexcorrespondingtoobjectextractionability,andevolvesintoanoptimalstructuringelementusingtheevolutionprinciplesofgeneticsearch.Experimentalresultsofroadextractionfromhighresolutionsatelliteimagesarepresentedtoillustratethemeritandfeasibilityoftheproposedmethod.

简介：Momentshavebeenusedinallsortsofobjectclassificationsystemsbasedonimage.TherearelotsofmomentsstudiedbymanyresearchersintheareaofobjectclassificationandoneofthemostpreferencemomentsistheZernikemoment.Inthispaper,theperformanceofobjectclassificationusingtheZernikemomenthasbeenexplored.Theclassifierbasedonneuralnetworkshasbeenusedinthisstudy.Theresultsindicatethebestperformanceinidentifyingtheaggregateisat91.4%withatenordersoftheZernikemoment.ThisencouragingresulthasshownthattheZernikemomentisasuitablemomenttobeusedasafeatureofobjectclassificationsystems.

简介：Theimpropercultivationofcommercialplantation,whichisamanagementactivitywiththepropertiesofeconomy,willleadtoenvironmentlossessuchaslanddegradation,soilerosion,pestsanddiseases.Environmentalcostiseasilyneglectedbyexecutives,however,itexistsactually.Thevariety,intersectingandoverlappingofenvironmentlossesdecidetheboundaryofevaluationandthenecessityofindicesonenvironmentalcost.Theresearchanalyzesthecharacteristicsanddisciplinariansofenvi...

简介：FractionalBrownianmotion,continuouseverywhereanddifferentiablenowhere,offersaconvenientmodelingforirregularnonstationarystochasticprocesseswithlong-termdependenciesandpowerlawbehaviorofspectrumoverwiderangesoffrequencies.Itshowshighcorrelationatcoasescaleandvariesslightlyatfinescale,whichissuitableforandsuccessfulindescribingandmodelingnaturalscenes.Ontheotherhand,man-madeobjectscanbeconstructivelywelldescribedbyusingasetofregularsimpleshapeprimitivessuchasline,cylinder,etc.andarefreeoffractal.Basedonthedifference,weprovideamethodtodiscriminateman-madeobjectsfromnaturalscenes.Experimentsareusedtodemonstratedthegoodefficiencyofthedevelopedtechnique.

简介：FractionalBrownianmotion,continuouseverywhereanddifferentiablenowhere,offersaconvenientmodelingforirregularnonstationarystochasticprocesseswithlong-termdependenciesandpowerlawbehaviorofspectrumoverwiderangesoffrequencies.Itshowshighcorrelationatcoarsescaleandvariesslightlyatfinescale,whichissuitableforandsuccessfulindescribingandmodelingnaturalscenes.Ontheotherhand,man-madeobjectscanbeconstructivelywelldescribedbyusingasetofregularsimpleshapeprimitivessuchasline,cylinder,etc.andarefreeoffractal.Basedonthedifference,amethodtodiscriminateman-madeobjectsfromnaturalscenesisprovided.Experimentsareusedtodemonstratethegoodefficiencyofdevelopedtechnique.

简介：ThereconstructionandsubsequentparticleidentificationisachallengeinacomplexandahighluminosityenvironmentsuchasthoseexpectedintheATLASdetectorattheLHC.TheATLASsoftwarehaschosentheobjectorientedparadigmandhasrecentlymigratedmuchofitssoftwarecomponentsdevelopedearlierusingproceduralprogramminglanguages.Thenewsoftware,whichemphasizesontheseparationbetweenalgorthmsanddataobjects,hasbeensuccessfullyintegratedinthebroaderATLASframework.Wewillpresentastatusreportofthereconstructionsoftwaresummarizingtheexperiencesgainedinthemigrationofseveralsoftwarecomponents.Wewillexaminesomeofthecomponentsofthecalorimetersoftwaredesign,whichincludesimulationofreal-timedetectoreffectsandonlineenvironment,andstrategiesdeployedforidentificationofparticles.

简介：ＯｐｔｉｃａｌＦｉｂｅｒＡｐｐｌｉｃａｔｉｏｎｔｏｔｈｅＯｂｊｅｃｔＲｅｃｏｇｎｉｔｉｏｎｏｆＤｅｘｔｅｒｏｕｓＨａｎｄ①ＸＩＥＸｕｑｉａｎｇ，ＷＡＮＧＳｈｏｕｙｕ，ＷＡＮＧＨｏｎｇｒｕｉ（ＹａｎｓｈａｎＵｎｉｖｅｒｓｉｔｙ，Ｑｉｎｈｕａｎｇｄａｏ０...

简介：VirtualassemblyisaVirtualReality(VR)basedengineeringapplicationwhichallowsengineerstoevaluate,analyze,andplantheassemblyofmechanicalsystems,Tomodelthevirtualassemblyprocess,newmethodologymustbeapplied.Basedontheideathatthevirtualassemblysystemisaneventdrivensystem,theinteractivebehaviorandinformationmodelisproposedtodescribethedynamicprocessofvirtualassembly.Definitionoftheobjectorientedmodelofvirtualassemblyisputforward.

简介：Byusingasphericalwaveasthereferencewave,werecordedthein-linephase-shiftingdigitalhologramofthe25thelementofChinesestandardNo.3resolutiontestpattern,andgavethecorrespondingnumericalreconstructedresults.Someproblemsconcerningwiththedigitalhologramrecordingandreconstructionofthediffractiveobjectatashortdistancearediscussed.Theexperimentalresultshowsthattheresolutionofthereconstructedimageisbetterthan10μm,whichisthelimitbyusingthisexperimentalarrangement.

简介：这份报纸论述一个粒子有联机特征选择机制的基于过滤器的视觉追踪方法。在基于颜色的粒子过滤器算法，粒子的重量总是不正确地代表重要性，这目标基于粒子追踪过滤的原因可以收敛到目标的一个本地区域。在我们的建议视觉追踪的方法，在目标和背景之间的Bhattacharyya距离和本地辨别被用来定义粒子的重量，它能解决存在本地集中问题。实验证明建议方法能不仅在追踪过程的单个目标而且在追踪过程的多重类似的目标工作很好。