Digital Image Processing: A Remote Sensing PerspectivePerspectiveFW5560Lecture 7Lecture 7Radiometric CorrectionRadiometric correctionRadiometric correctionImproving the accuracy of surface spectral reflectance or emittance. Would like to obtain measurements of absolute flihh’fM idreflectance or emittance at the earth’s surface. Must consider contributions and subtraction imposed by atmospheric components.pAlso consider gains and bias of the sensors themselves.Relationship between DN and Spectral RadianceSlope = channel gainIntercept =DNIntercept = channel offsetL = spectral radiance (watts/cm2/sr)RMAXRMIN= spectral radiance of internal calibration lampsL = spectral radiance (watts/cm2/sr)Radiometric Correction StepsOutput values for malfunctioning detectors are replaced by DNs of adjacent detectors in the same bandjBand responses to the calibration bands are use to characterize the in-orbit response functions of the individual channelsthe inorbit response functions of the individual channels. Functions are updated every 11,520 scan linesHi t li ti d i f dHistogram equalization procedure is performedRadiometric LUTs are calculated to convert the DNs to units of absolute radianceUsed to measure difference insscenrediancein a quantitativeUsed to measure difference in sscenrediancein a quantitative fashion within and between scenseHistogram equalization- minimize residual line to line stripinggqpgDN frequency histogram for each sensor of each band, calculate the mean and std dev Convert this numbers to absolute unitsthe mean and std dev. Convert this numbers to absolute units using the original estimates of gain and offset.’(/)G’c,b= Gc.b(Sc,b/Sb)B’cb= Bcb+ Gcb(Mcb–(SbMb/Scb))c,bc,bc.b(c,b(bbc,b))Gc.band Bc,b= original estimates of gain and offset respectivelyM= mean associate the histogram for sensor c and band b inMc,b= mean associate the histogram for sensor c and band b in units of absolute radianceSc,b= std dev. of the histogramSfh dddii f bdSb= avg of the standard deviations for a bandAtmospheric CorrectionAtmospheric CorrectionOptical characteristics of the atmosphere are estimated either by using special features of the ground surface or by direct measurements of the atmospheric constituents or by using theoretical models Various quantities related to the atmospherictheoretical models. Various quantities related to the atmospheric correction can then be computed by the radiative transfer algorithms given the atmospheric optical properties. Second, the tl di b tdb i i dremotely sensed imagery can be corrected by inversion procedures that derive the surface reflectance.Single Image Normalization Using Dark Image AdjustmentSingle Image Normalization Using Dark Image AdjustmentATCOR3 is atmospheric correction and haze removal softwareATCOR3 is atmospheric correction and haze removal software used to correct changes in the spectral reflectance of materials on the earth's surface. ATCOR3 can be integrated with a DEM for thi tiii f dtiatmospheric correction in images of rugged terrain. Features:Produce sharp and brilliant satellite images Reduce the shadow effect in mountainous terrain Obtain better classification results by using real reflectanceObtain better classification results by using real reflectance values Compare multi-temporal and multi-sensoral values Cl lt l ddd d t h k i f t d tCalculate value added products such as sky view factor and cast shadow