简介：Thispaperpresentsahumandetectionsysteminavision-basedhospitalsurveillanceenvironment.Thesystemiscomposedofthreesubsystems,i.e.backgroundsegmentationsubsystem(BSS),humanfeatureextractionsubsystem(HFES),andhumanrecognitionsubsystem(HRS).ThecodebookbackgroundmodelisappliedintheBSS,thehistogramoforientedgradients(HOG)featuresareusedintheHFES,andthesupportvectormachine(SVM)classificationisemployedintheHRS.Bymeansoftheintegrationofthesesubsystems,thehumandetectioninavision-basedhospitalsurveillanceenvironmentisperformed.Experimentalresultsshowthattheproposedsystemcaneffectivelydetectmostofthepeopleinhospitalsurveillancevideosequences.

简介：Traditionalbackgroundmodelmethodsoftenrequirecomplicatedcomputations,andaresensitivetoilluminationandshadow.Inthispaper,weproposeablock-basedbackgroundmodelingmethod,anduseourproposedmethodtocombinecolorandtexturecharacteristics.Suppressionandrelaxationarethetwokeystrategiestoresistilluminationchangesandshadowdisturbance.Theproposedmethodisquiteefficientandiscapableofresistingilluminationchanges.Experimentalresultsshowthatourmethodissuitableforreal-wordscenesandreal-timeapplications.

简介：因为在哪个录像运动需要被追踪，环境广泛地基于申请被改变并且在追踪的系统的设计和表演上提出几限制，录像追踪是一个复杂问题。没有彻底地分析环境强加的这些不同限制的效果，被用来评估并且比较录像运动追踪算法的当前的数据集使用一项累积性能措施。但是它需要作为参数分析这些限制。这份报纸的目的是识别这些参数并且定义量的措施让这些参数为运动追踪比较录像数据集。

简介：Super-Resolution(SR)techniquemeanstoreconstructHigh-Resolution(HR)imagesfromasequenceofLow-Resolution(LR)observations,whichhasbeenagreatfocusforcompressedvideo.BasedonthetheoryofProjectionOntoConvexSet(POCS),thispaperconstructsQuantizationConstraintSet(QCS)usingthequantizationinformationextractedfromthevideobitstream.BycombiningthestatisticalpropertiesofimageandtheHumanVisualSystem(HVS),anovelAdaptiveQuantizationConstraintSet(AQCS)isproposed.SimulationresultsshowthatAQCS-basedSRal-gorithmconvergesatafastrateandobtainsbetterperformanceinbothobjectiveandsubjectivequality,whichisapplicableforcompressedvideo.

简介：ThisletterproposesaratecontrolalgorithmforH.264videoencoder,whichisbasedonblockactivityandbufferstate.ExperimentalresultsindicatethatithasanexcellentperformancebyprovidingmuchaccuratebitrateandbettercodingefficiencycomparedwithH.264.ThecomputationalcomplexityofthealgorithmisreducedbyadoptinganovelblockactivitydescriptionmethodusingtheSumofAbsoluteDifference(SAD)of16×16mode,anditsrobustnessisenhancedbyintroducingafeedbackcircuitatframelayer.

简介：Thisletterproposesanefficientwavelet-basedFineGranularityScalable(FGS)codingscheme,wherethebaselayerisencodedwithanewlydesignedwavelet-basedcoder,andtheenhancementlayerisencodedwithProgressiveFineGranularityScalable(PFGS)coding.Thisalgorithminvolvesmulti-framemotioncompensation,rate-distortionoptimizingstrategywithLagrangiancostfunctionandcontext-basedadaptivearithmeticcoding.Inordertoimproveefficiencyoftheenhancementlayercoding,animprovedmotionestimationschemethatusesbothinformationfromthebaselayerandtheenhancementlayerisalsoproposedinthisletter.Thewavelet-basedcodersignificantlyimprovesthecodingefficiencyofthebaselayercomparedwithMPEG-4ASP(AdvancedSimpleProfile)andH.26LTML9.ThePFGScodingisasignificantimprovementoverMPEG-4FGScodingattheenhancementlayer.Experimentsshowthatsinglelayercodingefficiencygainoftheproposedschemeisabout2.0-3.0dBand0.3-1.0dBhigherthanthatofMPEG-4ASPandH.26LTML9,respectively.Theoverallcodingefficiencygainoftheproposedschemeisabout4.0-5.0dBhigherthanthatofMPEG-4FGS.

简介：PresentedisaschemeofanembeddedvideoremotemonitoringsystembasedonTMS320DM642.UsingDM642asthedataprocessingcore,theremotemonitoringsystemiscomposedofvideoacquisitionmodule,videoprocessingmoduleandcommunicationmodule,whichgivesanimplementationofclass/minidrivermoduleinDSP/BIOSintegrateddevelopingenvironmentandalsoacommontaskmoduleinapplicationlayerisachieved.Thesystemrealizestheentirefunctionsoftheanalogvideosignalacquisition,H.264videocodingandInternettransmission.Itprovidesthegeneralconnectionforthefuturedevelopmentandhasgoodflexibilityandextendibility.Thesystemusesamodulardesignandoveralldevelopmentofprogrammingmethodstoimprovetheefficiencyofsystemdevelopment.

简介：Inthispaperweproposeathreedimensionalmultiplierlessdiscretecosinetransform(DCT)withliftingschemecalled3D-binDCT.Basedon3D-binDCT,anovelvideocodingalgorithmwithoutmotionestimation/compensationisproposed.Itusesthe3D-binDCTtoexploitspatialortemporalredundancy.ThecomputationofbinDCTonlyneedsshiftandadditionoperations,thusthecomputationalcomplexityisminimized.DCcoefficientprediction,modifiedscanmodeandarithmeticcodingtechniquesarealsoadopted.Extensivesimulationresultsshowthattheproposedcodingschemeprovideshighercodingefficiencyandimprovesvisualquality,anditiseasytoberealizedbysoftwareandhardware.

简介：Movingobjectextractionandclassificationareimportantproblemsinautomatedvideosurveillancesystems.Abackgroundmodelbasedonregionsegmentationisproposed.AnadaptivesingleGaussianbackgroundmodelisusedinthestableregionwithgradualchanges,andanonparametricmodelisusedinthevariableregionwithjumpingchanges.Ageneralizedagglomerativeschemeisusedtomergethepixelsinthevariableregionandfillinthesmallinterspaces.Atwo-thresholdsequentialalgorithmicschemeisusedtogroupthebackgroundsamplesofthevariableregionintodistinctGaussiandistributionstoacceleratethekerneldensitycomputationspeedofthenonparametricmodel.Inthefeature-basedobjectclassificationphase,thesurveillancesceneisfirstpartitionedaccordingtotheroadboundariesofdifferenttrafficdirectionsandthenre-segmentedaccordingtotheirscenelocalities.Themethodimprovesthediscriminabilityofthefeaturesineachpartition.AdaBoostmethodisappliedtoevaluatetherelativeimportanceofthefeaturesineachpartitionrespectivelyanddistinguishwhetheranobjectisavehicle,asinglehuman,ahumangroup,orabike.Experimentalresultsshowthattheproposedmethodachieveshigherperformanceincomparisonwiththeexistingmethod.

简介：Toaimathighercodingefficiencyformultiviewvideocoding,themultiviewvideowithamodifiedhighefficiencyvideocoding(MV-HEVC)codecisproposedtoencodethedependentviews.However,thecomputationalcomplexityofMV-HEVCencoderisalsoincreasedsignificantlysinceMV-HEVCinheritsallcomputationalcomplexityofHEVC.ThispaperpresentsanefficientalgorithmforreducingthehighcomputationalcomplexityofMV-HEVCbyfastdecidingthecodingunitduringtheencodingprocess.Inourproposal,thedepthinformationofthelargestcodingunits(LCUs)fromindependentviewandneighboringLCUsisanalyzedfirst.Afterwards,theanalyzedresultsareusedtoearlydeterminethedepthfordependentviewandthusachievecomputationalcomplexityreduction.Furthermore,apredictionunit(PU)decisionstrategyisalsoproposedtomaintainthevideoquality.Experimentalresultsdemonstratethatouralgorithmcanachieve57%timesavingonaverage,whilemaintaininggoodvideoqualityandbit-rateperformancecomparedwithHTM8.0.

简介：Ahybridpilotsassistedchannelestimationalgorithmformultipleinputmultipleoutput(MIMO)orthogonalfrequencydivisionmultiplexing(OFDM)systemsunderlowsignal-to-noiseratio(SNR)andarbitraryDopplerspreadscenariosisproposed.Motivatedbythedissatisfactoryperformanceoftheoptimalpilots(OPs)designedunderstaticchannelsovermultipleOFDMsymbolsimposedbyfastfadingchannels,theproposedschemefirstassumesthatthevirtualpilottonessuperimposedatdatalocationsoverspecificsubcarriersaretransmittedfromallantennas,thenthevirtualreceivedpilotsignalsatthecorrespondinglocationscanbeobtainedbymakingfulluseofthetimeandfrequencydomaincorrelationsofthefrequencyresponsesofthetimevaryingdispersivefadingchannelsandthereceivedsignalsatpilotsubcarriers,finallythechannelparametersarederivedfromthecombinationoftherealandvirtualreceivedpilotsignalsoveroneOFDMsymbolbasedonleastsquare(LS)criterion.SimulationresultsillustratethattheproposedmethodisinsensitivetoDopplerspreadandcaneffectivelyamelioratethemeansquareerror(MSE)floorinherenttothepreviousmethod,meanwhileitsperformanceoutmatchesthatofOPsatlowSNRregionunderstaticchannels.

简介：Videoobjectextractionisakeytechnologyincontent-basedvideocoding.AnovelvideoobjectextractingalgorithmbytwoDimensional(2-D)mesh-basedmotionanalysisisproposedinthispaper.Firstly,a2-Dmeshfittingtheoriginalframeimageisobtainedviafeaturedetectionalgorithm.Then,higherorderstatisticsmotionanalysisisappliedonthe2-Dmeshrepresentationtogetaninitialmotiondetectionmask.Afterpost-processing,thefinalsegmentingmaskisquicklyobtained.Andhencethevideoobjectiseffectivelyextracted.Experimentalresultsshowthattheproposedalgorithmcombinesthemeritsofmesh-basedsegmentingalgorithmsandpixel-basedsegmentingalgorithms,andherebyachievessatisfactorysubjectiveandobjectiveperformancewhiledramaticallyincreasingthesegmentingspeed.

简介：InordertoimprovethevideotransmissionperformanceinMultiple-InputMultiple-OutputOrthogonalFrequencyDivisionMultiplexing(MIMO-OFDM)system,anewscheme,whichintegratesMultipleDescriptionCoding(MDC),LowDensityParityCheck(LDPC)codingandhybridspacetimecoding,isproposedinthisletter.Inparticular,ahybridspacetimecodingalgorithmiscombinedwithLDPCcodestoperformUnequalErrorProtection(UEP)ofMDCencodedvideostreams.ComparingwiththeUEPtransmissionwithonlyLDPCcodes,theproposedschemeachievesmorethan1dBgainintermsofSignaltoNoiseRatio(SNR)whenthePeakSignaltoNoiseRatio(PSNR)ofreconstructedvideoisabove30dB.

简介：Thewirelesssensornetwork(WSN)consistsofsensornodesthatinteractwitheachothertocollectivelymonitorenvironmentalorphysicalconditionsatdifferentlocationsfortheintendedusers.Oneofitspotentialdeploymentsisintheformofsmarthomeandambientassistedliving(SHAAL)tomeasurepatientsorelderlyphysiologicalsignals,controlhomeappliances,andmonitorhome.ThispaperfocusesonthedevelopmentofawirelesssensornodeplatformforSHAALapplicationoverWSNwhichcomplieswiththeIEEE802.15.4standardandoperatesin2.4GHzISM(industrial,scientific,andmedical)band.TheinitialstageofSHAALapplicationdevelopmentisthedesignofthewirelesssensornodenamedTelGmote.ThemainfeaturesofTelGmotecontributingtothegreencommunicationsincludelowpowerconsumption,wearable,flexible,user-friendly,andsmallsizes.Itisthenembeddedwithaself-builtoperatingsystemnamedWiseOStosupportcustomizedoperation.Thenodecanachieveapacketreceptionrate(PRR)above80%foradistanceofupto8m.ThedesignedTelGmoteisalsocomparablewiththeexistingwirelesssensornodesavailableinthemarket.

简介：Thepaperfirstdiscussesshortcomingsofclassicaladjacent-framedifference.Secondly,basedontheimageenergyandhighorderstatistic(HOS)theory,backgroundreconstructionconstraintsaresetup.Underthehelpofblock-processingtechnology,backgroundisreconstructedquickly.Finally,backgrounddifferenceisusedtodetectmotionregionsinsteadofadjacentframedifference.TheDSPbasedplatformtestsindicatethebackgroundcanberecoveredlosslesslyinaboutonesecond,andmovingregionsarenotinfluencedbymovingtargetspeeds.Thealgorithmhasimportantusagebothintheoryandapplications.