简介：匹配算法与的适应加权的立体声多，水平和双向动态编程基于地面控制点(GCP)被介绍。到没有失去匹配的精确的减少时间复杂性，使用一多铺平搜索计划，粗糙的匹配在典型不同空间图象被处理，当好匹配在不同偏移量空间被处理时，想象。在上面的水平，GCP被强制相互的限制和阀值限制的提高的容量的反复的算法获得。在高度可靠的GCP的管理下面，双向动态编程框架被采用在优化路径解决矛盾。在底层，到运用时间的还原剂，不同偏移量空间被建议高效地完成稠密的不同图象。另外，适应双支持重量策略被介绍聚集匹配的费用，它考虑光度计、几何的信息。进一步，processing以后算法能改善在有深度断绝的区域的不同结果并且由用双阀值算法的吸藏相关，在错过立体声的地方，信息从包围区域被代替。表明算法的有效性，我们在场四的二组试验性的结果广泛地使用了标准立体声数据集合包括有另外的方法的性能和比较的讨论，它证明算法有快速度不仅，而且显著地改进整体的优化的效率。

简介：Inthispaper,aneffectiveandrobustactivespeechdetectionmethodisproposedbasedonthe1/fprocesstechniqueforsignalsundernon-stationarynoisyenvironments.TheGaussian1/fprocess,amathematicalmodelforstatisticallyself-similarradomprocessesbasedonfractals,isselectedtomodelthespeechandthebackgroundnoise.AnoptimalBayesiantwo-classclassifierisdevelopedtodiscriminatethembytheir1/fwaveletcoefficientswithKarhunen-Loeve-typeproperties.Multipletemplatesaretrainedforthespeechsignal,andtheparametersofthebackgroundnoisecanbedynamicallyadaptedinruntimetomodelthevariationofboththespeechandthenoise.Inourexperiments,a10-minutelongspeechwithdifferenttypesofnoisesrangingfrom20dBto5dBistestedusingthisnewdetectionmethod.Ahighperformancewithover90%detectionaccuracyisachievedwhenaverageSNRisabout10dB.

简介：

简介：

简介：InordertoimprovetheperformanceofdeceptiondetectionbasedonChinesespeechsignals,amethodofsparsedecompositiononspectralfeatureisproposed.First,thewaveletpackettransformisappliedtodividethespeechsignalintomultiplesub-bands.Bandcepstralfeaturesofwaveletpacketsareobtainedbyoperatingthediscretecosinetransformonlogarithmicenergyofeachsub-band.ThecepstralfeatureisgeneratedbycombingMelFrequencyCepstralCoefficientandWaveletPacketBandCepstralCoefficient.Second,K-singularvaluedecompositionalgorithmisemployedtoachievethetrainingofanover-completemixturedictionarybasedonboththetruthanddeceptivefeaturesets,andanorthogonalmatchingpursuitalgorithmisusedforsparsecodingaccordingtothemixturedictionarytogetsparsefeature.Finally,recognitionexperimentsareperformedwithvariousclassifiedmodules.Experimentalresultsshowthatthesparsedecompositionmethodhasbetterperformancecompariedwithconventionaldimensionreducedmethods.Therecognitionaccuracyofthemethodproposedinthispaperis78.34%,whichishigherthanmethodsusingotherfeatures,improvingtherecognitionabilityofdeceptiondetectionsystemsignificantly.

简介：在从多时间的SAR图象产生的一幅木头比率差别图象考虑空间上下文的信息的一个无指导的变化察觉方法被建议。提起的AMarkov随机(MRF)模特儿特别地被雇用在附近的象素之中利用紧张层次的统计空间关联。在在木头比率差别图象的象素和混合Gaussian分发的独立的假设下面，一个随机、反复的EM-MPMchange-detection算法基于一个MRF模型被开发。EM-MPM算法基于以后的marginals(MPM)的maximiser为图象分割和期望最大值(他们)的算法为以一个完全自动的方法的参数评价的算法。在多时间的ERS-2SAR图象上获得的实验结果显示出建议方法的有效性。

简介：Wepresentaninexpensivetechniquetoobtainathree-dimensional(3D)millimeterwave(MMW)andterahertz(THz)imageusingupconversion.InthisworkwedescribeanddemonstrateamethodforupconversionofMMW/THzradiationtothevisualbandusingaveryinexpensiveminiatureglowdischargedetector(GDD)andasiliconphotodetector.WepresentMMW/THzupconversionimagesbasedonmeasuringthevisuallightemittingfromtheGDDratherthanitselectricalcurrent.Theresultsshowbetterresponsetimeandbettersensitivitycomparedtotheelectronicdetectionperformedpreviously.Furthermore,inthisworkweperformfrequencymodulationcontinuouswave(FMCW)radardetectionbasedonthismethodusingaGDDlamp,withaphotodetectortomeasureGDDlightemission.ByusingFMCWdetection,therangeinadditiontotheintensityateachpixelcanbeobtained,thusyieldingthe3Dimage.TheGDDactsasaheterodynemixernotonlyelectronicallybutalsooptically.Thesuggested3DupconversiontechniqueusingtheGDDissimpleandinexpensiveandhasbetterperformancecomparedtootherMMW/THzimagingsystemssuggestedintheliterature.Thismethodprovidesminimumdetectablesignalpowerthatisabout6ordersofmagnitudebetterthansimilarplasmasystemsduetotheverylargeinternalsignalgainderivingfromthemuchsmallerelectrodeseparationandresultinginmuchhigherplasmaelectricfield.

简介：这篇文章学习交流电领域测量(ACFM)的申请在缺点察觉为的方法在水下结构。ACFM系统的数字模型在海水环境为结构表面缺点察觉被造。有限元素模拟被执行在背叛区域调查电磁的信号分发的规则和特征。关于模拟结果，在水下人工的裂缝察觉实验为ACFM系统被设计并且进行。ACFM系统能检测的实验结果表演裂开在在水下组织，察觉精确性比85%高。这能满足设计要求在水下结构缺点察觉。在这篇文章的结果能被使用为基于的ACFM的优化和开发建立技术基础在水下结构缺点察觉系统。

简介：与逐渐地变得的产品改变了并且可以证明改善了的遥感信息，如此的产品的科学应用依靠他们的优秀评价。在象NASA(国家航空学和空间管理)那样的运作的上下文信息生产基于MODIS(中等分辨率成像分光辐射函数)在观察系统(曙光女神)地和水卫星的土上的仪器，检测产品异例的有效方法，即，为了在地球进程在产品人工制品和真实变化之间区别，被监视，是极其重要的在刺关于潜在的unreliability帮助并且通知用户社区为异例察觉的一种技术，是知道疯(中部绝对背离从中部)在经由系列分析被描述的时间的MODIS陆地产品，它能处理intra--并且在由使用原来的数据和他们的一阶的差别的疯统计的数据的内部年度的变化。这个方法被显示柔韧并且越过主要陆地产品工作，包括NDVI，活跃的火，雪盖子，和表面反射，并且它到多学科的产品的适用性被期望。

简介：“Seminarforhydrocarbondetectionwithcompositegeophysical/geo-chemicaltechniques”,jointlyorganizedbyChinapetroleumExploration&ProductionCompanyandExplorationGeophysicalCommitteeofCGSandsupportedbytheCompositeGeophysical/geo-chemicalDepartementofOrientalGeophysicalCompanyandChinaExploration&DevelopmentResearchInstiute,

简介：Basedonacurrentfogdetectiontheory,amultibandthresholdmethodforMODISdatawasputforwardtodetectdaytimefogintheSouthChinaSea.ItusedBands1,2,18,20and31ofMODISdatatoseparatefogfromthecloudandtheseasurface.Thedigitaldetectionindexeswereasfollows.IfRB1<20%,RB2<20%andRB1>RB2,thepixelwasidentifiedtobetheseasurface.IfRB1>55%,RB2>55%andTB31<273K,thepixelwasidentifiedtobeamiddle-andhigh-levelcloud.IfIFC>20,thepixelwasclassifiedtobeseafog.ThemethodwasverifiedwithseafogdataobservedfromthecoastalregionofGuangdongduringJanuary-April2011.Outofthe13samplesofsatellitedetection,ninewereconsistentwiththesurfaceobservations,threewereidentifiedtobelow-levelthecloudaccordingtothesatellitedetectionbutfogaccordingtothesurfaceobservations,andonlyonesamplewasidentifiedtobetheoceansurfacebythesatellitedetectionbutfogbythesurfaceobservations.BecausetheMODISdatacannotpenetratethecloudorfog,themodelwasdesignedforasinglefieldofviewwhichhadonlyonelayerofcloudorfog.Itcanaccuratelydistinguishfogwhichisnotcoveredbythecloud,butitidentifiesfogascloudiftheformeriscoveredbyacloud.Generallyspeaking,themodeliseffectiveandfeasible.

简介：Amethodofdetectingdry,icyandwetroadsurfaceconditionsbasedonscanningdetectionofsinglewavelengthbackwardpowerisproposedinthisletter.Thedetectorisusedtoreceivethebackwardscatteredpowerwhichchangeswiththeincidenceangle.Therelationshipbetweenbackwardpowerandincidenceangleisusedtofindouttheeffectiveanglerangeanddistinguishmethod.Experimentandsimulationshowthatitisfeasibletoclassifythesethreeconditionswithinincidenceangleof5.3degree.

简介：许可证字符抽取的传统的方法不能满足识别精确性的要求，速度由录像显示车辆的察觉系统。因此，基于多尺度的边，察觉和一个字符分割算法基于Markov随机回答模型的一个牌照本地化方法被介绍。实验的结果证明方法基于阀值由差别操作员和特性分割基于边察觉与传统的本地化方法相对照产出更精确的许可证特性抽取。，精确性在差的状况下面在更好的照明下面从90%增加到94%，它增加超过5%。当二个改进算法被使用时，自动许可证识别的精确性和速度遇见系统“甚至在吵闹的情形或不平的照明下面的s要求。

简介：Thewaveletpackettransformisusedforthedamagedetectionofoffshoreplatforms.Whensomedamageoccurs,thedynamicresponseparametersofthestructurewillshiftsubtly.However,insomecases,thedynamicparameters,suchasdynamicresponse,arenotsensitive,anditisverydifficulttopredicttheexistenceofdamage.Thepresentpaperaimstodescribehowtofindsmalldamagebytheuseofwaveletpackettransform.Asthewaveletpackettransformcanbeusedtoquicklyfindthesingularityoftheresponsesignalondifferentscales,theaccelerationsignalofadamagedoffshoreplatforminthetimedomainistransformedthroughthewaveletpacket.ExperimentalresultsshowthattheDaubechies4wavelettransformcanbeusedtodetectdamage.

简介：

简介：Combiningtheinhibitedaptazymeandmolecularbeacon(MB),wedevelopedaversatilesensingstrategyforamplifieddetectionofadenosine.Inthisstrategy,theadenosineaptamerlinkstothe8-17DNAzymetoformanaptazyme.Ashortsequence,denotedasinhibitor,isdesignedtoformaduplexspanningtheaptamer–DNAzymejunction,whichblocksthecatalyticfunctionoftheDNAzyme.Onlyinthepresenceoftargetadenosine,theaptamerbindstoadenosine,thustheinhibitordissociatesfromtheaptamerportionoftheaptazymeandcannolongerformthestableduplexrequiredtoinhibitthecatalyticactivityoftheaptazyme.ThereleasedDNAzymedomainwillhybridizetotheMBandcatalyzethecleavageinthepresenceofZn2+,makingthefluorophoreseparatefromthequencherandresultinginfluorescencesignal.Theresultsshowedthatthedetectionmethodhasadynamicrangefrom10nmol/Lto1nmol/L,withadetectionlimitof10nmol/L.

简介：Inthedesignproblemoflowearthorbit(LEO)reconnaissancesatelliteconstellation,optimizationofcoverageperformanceisthedesigngoalinmostcurrentmethods.However,intheusingprocess,theuseronlyconcernswiththedetectioncapabilitiesratherthancoverageperformance.Toestablishtherelationshipbetweenthesetwoaspects,thereconnaissanceprocessesofnormalstochastictargetsareconsideredandthemathematicmodelsofdetectionprocessesarebuilt.Theindicatorsofcoverageperformanceareusedtoevaluatethedetectionprobabilityandexpectationofdetectiontimedelay,whichareimportantfactorsinreconnaissanceconstellationestimationviewedfrommilitaryintelligencediscipline.TheconclusionsconfirmedbythefinalsimulationwillbeusefulinLEOreconnaissanceconstellationdesign,optimizationandevaluation.

简介：Duringthelastdecade,spatio-temporaldatabaseshavebecomeincreasinglyimportantinmanyapplicationssuchasgeographicinformationsystems(GIS)andengineeringinformationsystems.Thispaperdiscussesthedesignandimplementationofageocomputingplatformforthedevelopmentoflocation-basedservices(LBS)focusingonmobilemapping.Duringtheanalysis,design,andimplementationofthegeocomputingplatform,aneffectivemethodisproposedforthereal-timeprocessingofgeographicinformationacquiredbyacameraattachedtoapersonaldigitalassistant(PDA).Thismethodcombineslocationinformationgivenbytheglobalpositionsystem(GPS)withman’sabilitytorecognizethelocationofobjectsandtheirgeographicalrelationshiptoimproveobjectmapping.

简介：

简介：AbstractDrug resistance via drug-resistant mutations in the human immunodeficiency virus (HIV) genome is the primary cause of antiviral therapy failure. Consequently, HIV drug resistance genotyping has become a critical approach in HIV prevention and control. Compared to the Sanger sequencing technology, high-throughput sequencing (HTS) technology has superior sensitivity and timeliness, with strong detection capabilities for low-frequency mutations. With the continued advancement of HTS technologies, their prominence in HIV drug resistance detection techniques has increased accordingly. This article will review the latest developments in HTS technology and its applications in HIV drug resistance testing.