简介:Wepresentapseudo-inverseghostimaging(PGI)techniquewhichcandramaticallyenhancethespatialtransverseresolutionofpseudo-thermalghostimaging(GI).IncomparisonwithconventionalGI,PGIcanbreakthelimitationontheimagingresolutionimposedbythespeckle’stransversesizeontheobjectplaneandalsoenablesthereconstructionofanN-pixelimagefrommuchlessthanNmeasurements.Thisfeaturealsoallowshigh-resolutionimagingofgray-scaleobjects.Experimentalandnumericaldataassessingtheperformanceofthetechniquearepresented.
简介:Anideaofestimatingthedirectsequencespreadspectrum(DSSS)signalpseudo-noise(PN)sequenceispresented.WithouttheapriorityknowledgeabouttheDSSSsignalinthenon-cooperationcondition,weproposeaself-organizingfeaturemap(SOFM)neuralnetworkalgorithmtodetectandidentifythePNsequence.Anon-supervisedlearningalgorithmisproposedaccordingtheKohonenruleinSOFM.TheblindalgorithmcanalsoestimatethePNsequenceinalowsignal-to-noise(SNR)andcomputersimulationdemonstratesthatthealgorithmiseffective.Comparedwiththetraditionalcorrelationalgorithmbasedonslip-correlation,theproposedalgorithm’sbiterrorrate(BER)andcomplexityarelower.