Thedatasignalobtainedbyusingthenear-infraredspectroscopydetectorincludestheinformationofthecomponenttobetested,includingthenoiseofvariousinstruments,suchashigh-frequencyrandomnoise,baselinedrift,spuriousinformation,backgroundofthesample.Therefore,beforedataanalysis,thespecificsignalmeasurementandsamplesystemshouldbeprocessedreasonably,andtheinfluenceofvariousnon-targetfactorsonthedetectionsignalinformationshouldbeweakenedoreveneliminated,whichlaysafoundationforestablishingastableandreliablemathematicalmodel.Co妹妹onlyuseddatapreprocessingmethodsincludedatanormalizationprocessing(meancentering,normalization,standardnormaltransformation,etc.),highfrequencynoisefiltering(convolutionsmoothing,Fouriertransform,wavelettransform,etc.),differentialderivationofsignals,andBaselinecorrection,etc.
1