预览加载中,请您耐心等待几秒...
1/3
2/3
3/3

在线预览结束,喜欢就下载吧,查找使用更方便

如果您无法下载资料,请参考说明:

1、部分资料下载需要金币,请确保您的账户上有足够的金币

2、已购买过的文档,再次下载不重复扣费

3、资料包下载后请先用软件解压,在使用对应软件打开

基于MFCC特征和隐马尔可夫模型的咳嗽信号自动识别 Abstract: Coughingisacommonsymptominavarietyofrespiratorydiseases.Withtheadvancementoftechnology,coughsignalrecognitiontechnologyhasbeenwidelyusedinmedicalandpublichealthfields.ThisarticleputsforwardthecoughsignalrecognitionmodelbasedonMFCCfeaturesandhiddenMarkovmodel.Firstly,thecoughsignalispreprocessedtoobtaintheMFCCfeaturevector.Secondly,thetrainingsetandtestsetareestablished.Finally,thehiddenMarkovmodelisusedforcoughrecognition.Theexperimentalresultsshowthattheproposedmethodhasgoodrecognitionaccuracyandreal-timeperformance,andcanbeappliedtotherecognitionofcoughsignalsinpublichealthandmedicalfields. Introduction: Coughisacommonsymptominrespiratorydiseasessuchasasthma,pneumonia,andtuberculosis.Coughingcanhelpcleartherespiratorytractofsecretionsandforeignmatter.However,coughingcanalsobeasignofanunderlyingmedicalcondition.Inaddition,coughingcanspreadinfectiousdiseases.Therefore,coughsignalrecognitiontechnologyhasbecomeanimportantmeansofmonitoringrespiratoryhealthandcontrollingthespreadofrespiratorydiseases. MFCC(Mel-FrequencyCepstralCoefficients)featureisacommonlyusedaudiosignalfeature.Ithastheadvantagesofgoodtime-frequencyanalysisability,highdatacompressionrate,andeffectivenoisereduction.HiddenMarkovmodel(HMM)isaprobabilisticmodelwidelyusedinsignalrecognition.HMMisbasedontheMarkovprocessandcaneffectivelydealwiththeproblemofsignalsequencemodeling. CoughsignalrecognitionbasedonMFCCfeaturesandHMMhasbecomearesearchhotspotinrecentyears.Inthispaper,acoughsignalrecognitionmodelbasedonMFCCfeaturesandHMMisproposed.Theproposedmodelcanachievehighrecognitionaccuracyandreal-timeperformance,andhaspracticalapplicationvalueinthefieldofpublichealthandmedicalcare. Methodology: PreprocessingofCoughSignal: Thecoughsignalisfirstcollectedbyamicrophone,andthenthenoiseisremovedbywaveletthresholddenoising.Next,thecoughsignalispreprocessedbyframing,windowing,andfastFouriertransform.Thesignalisdividedinto25millisecondframeswitha10millisecondoverlap.TheHammingwindowi