基于暗原色的航空遥感影像薄雾去除与改进.docx

基于暗原色的航空遥感影像薄雾去除与改进
Title:FogRemovalandEnhancementofAirborneRemoteSensingImagesBasedonDarkChannelPrior
Abstract:
Airborneremotesensinghasbecomeaninvaluabletoolforvariousapplications,includinglandcovermapping,environmentalmonitoring,anddisastermanagement.However,onemajorchallengefacedintheanalysisandinterpretationoftheacquiredimagesisthepresenceoffog,whichcansignificantlydegradethevisualqualityandlimittheaccuracyofsubsequentimageprocessingtasks.Thispaperexplorestheapplicationofthedarkchannelprioralgorithmforfogremovalandenhancementofairborneremotesensingimages.Thealgorithmtakesadvantageofthestatisticalpropertiesofnaturalimagestoestimatetheatmosphericlightandtransmissionmap,allowingforeffectiverestorationofthefoggyscenes.
Introduction:
Airborneremotesensingplatforms,suchassatellitesandaerialdrones,offerauniqueperspectiveforcapturinghigh-resolutionimagesoftheEarth'ssurface.However,inmanyregions,atmosphericphenomena,suchasfog,haze,andsmog,canobstructtheclarityoftheacquiredimages.Amongtheseatmosphericconditions,fogisaparticularchallengeduetoitsdenseandspatiallyvaryingnature.Removingfogfromairborneremotesensingimagesisvitaltoimprovevisibility,enhanceinterpretability,andfacilitatesubsequentimageanalysistasks.
DarkChannelPrior:
Thedarkchannelprioralgorithmiswidelyusedforhazeorfogremovalinnaturalimages.Itexploitstheobservationthatinmostoutdoorimages,theintensityofatleastonecolorchannelinnon-foggyregionsisclosetozeroduetothescatteringoflightbytheairmoleculesandthepresenceofaerosols.Byidentifyingthepixelswiththelow-intensityvaluesacrossthecolorchannels,anestimationoftheatmosphericlightcomponentcanbeobtained.Additionally,thealgorithmcalculatesthetransmissionmap,whichrepresentstheratioofthesceneradiancetotheatmosphericlightateachpixel.Thistransmissionmapallowsfortherestorationoftheoriginalsceneinformationbyattenuatingtheeffectsoffog.
AdaptationforAirborneRemoteSensingImages:
Toadaptthedarkchannelprioralgorithmforairborneremotesensingimages,certainconsiderationsneedtobetakenintoa