摘要
Atpresent,boththepointsourceandtheimagingpolarizationnavigationdevicesonlycanoutputtheangleinformation,whichmeansthatthevelocityinformationofthecarriercannotbeextractedfromthepolarizationfieldpatterndirectly.Opticalflowisanimage-basedmethodforcalculatingthevelocityofpixelpointmovementinanimage.However,forordinaryopticalflow,thedifferenceinpixelvalueaswellasthecalculationaccuracycanbereducedinweaklight.Polarizationimagingtechnologyhastheabilitytoimproveboththedetectionaccuracyandtherecognitionprobabilityofthetargetbecauseitcanacquiretheextrapolarizationmulti-dimensionalinformationoftargetradiationorreflection.Inthispaper,combiningthepolarizationimagingtechniquewiththetraditionalopticalflowalgorithm,apolarizationopticalflowalgorithmisproposed,anditisverifiedthatthepolarizedopticalflowalgorithmhasgoodadaptationinweaklightandcanimprovetheapplicationrangeofpolarizationnavigationsensors.Thisresearchlaysthefoundationfordayandnightall-weatherpolarizationnavigationapplicationsinfuture.
出版日期
2018年03月13日（中国期刊网平台首次上网日期，不代表论文的发表时间）
