1、Fundamental for SAR Remote SensingPassive remote sensing Active remote sensingPassive systemOptical sensorsActive systemradar sensorsRADAR=RAdio Detection And Ranging,radar=a ranging or distance measuring device with a transmitter,a receiver,an antenna,and an electronics system to process and record
2、 the data.Transmitter generates successive short bursts(or pulses of microwave(A)at regular intervals which are focused by the antenna into a beam(B).The antenna receives a portion of the transmitted energy reflected(or backscattered)from various objects within the illuminated beam(C).Radar and elec
3、tromagnetic spectrumRadar signal:amplitude and phaseSAR signal is a complex natureSAR=Amplitude+phaseSpeckle appears as a grainy salt and pepper texture in an image.This is caused by random constructive and destructive interference from the multiple scattering returns that will occur within each res
4、olution cell.Homogeneous targetConstructive interferenceDestructive interferenceVarious level of interference(between constructive and destructive)Signal-Surface interactions and back-scatteringMicrowave interaction with natural surfaces is a function of:SAR configuration incidence angle frequency p
5、olarisation Surface s physical and electrical properties roughness soil moisture vegetation parameters Surface s geometry Local and large scaleSignal-Surface interactions:Rayleigh criterion Specular reflection (smooth surface)h /8cos Diffuse reflection (rough surface)hradar response due to backscatt
6、ered signal2 Penetration in a dry soil(2):response depends upon subsurface layer3-Smooth surface or a super-saturated soil(3):Specular reflection=low radar responseNote:influence of radar wavelength and incidence angleReflection off forest canopy(1)Penetration into foliage(2)(volume scattering)Multi
7、ple reflections on trees and soil(3)2311-Low incidence angle and short wavelengths(1)=:no penetration into vegetation2-Strong incidence and large wavelengths(2):=penetration into vegetation=information on moisture,soil and/or water3-Specific temporal behavior of certain crop types (e.g.rice)21Radar
8、signal-surface interactionsFree(liquid)water=smooth surfaceUrban area=rough surfacePloughed fields=rough surfaceCrop at early stage=smooth surfacedarkbrightdarkbrightSAR:Geometry of IlluminationRadarAzimuth DirectionSwathHGround TrackRangeSynthetic Aperture PrincipleTimeABCEDPDopplerFrequency(of P)S
9、atellitePathGroundSAR spatial resolutionfauchehi Antenne du radar Range resolution ihLaadraAzimuth resolution:determined by the angular width of the radiated microwave beam and the slant range distance Radar Image distortion:Slant-range scale distortion Radar=measure the distance to features in slan
10、t-range rather than the true horizontal distance along the ground.a varying image scale,moving from near to far range Targets A1 and B1:same size on the ground,Different apparent dimensions in slant range(A2 and B2)Targets in the near range to appear compressed relative to the far range.Using trigon
11、ometry,ground-range distance can be calculated from the slant-range distance and platform altitude to convert to the proper ground-range format AB CDImage brightness Effect of topography:-Shadows-Foreshortening-Layover (extreme case of foreshortening,topography is inverted)=SAR use limitations in re
12、lief areaShadowUse of DEMto correctA=Smaller=foreshorteningRADAR Geometric Distortions:Foreshortening-Shadowing-LayoverInfluence of systems configuration:polarizationPolarization effectsPolarization Refers to the orientation of the electric field Transmit microwave radiation either horizontally pola
13、rized(H)or vertically polarized(V).Antenna receives either the horizontally or vertically polarized backscattered energy,and some radars can receive both.Four combinations:HH-for horizontal transmit and horizontal receive,VV-for vertical transmit and vertical receive,HV-for horizontal transmit and v
14、ertical receive,VH-for vertical transmit and horizontal receiveInfluence of systems configuration:polarization and wavelengthPolarization effectsCLWavelength effectsComparison of optical and SAR data OPTICALBiological Sensitivity(ex:photosynthesis)High ResolutionSpectral Resolution(several bands)RAD
15、ARPhysical Sensitivity(ex:roughness and moisture)High ResolutionAll weather acquisition(cloud penetration)Image processing techniques of SAR dataGeometric correction and geocodingCalibrationSpeckle filteringCoherence imagesClassificationAmplitude and phase SAR informationQuantitative analysis of SAR
16、 dataPrerequisite:Calibration(documented by ESA)Measurement accuracy ERS signal stability(+/-0.2 dB)Observation scale0 =f(roughness)*f(soil moisture)*f(vegetation parameter)1 date(1 equation,several unknown parameters?)multi-date,synergy with other sensors,SIGPoor scientific knowledge“SAR signal int
17、eraction with natural surfaces”refNiiiKDNNsinsin11log10102SAR data:interesting information Difficulty to use or extractSpeckle affecting photo-interpretation introducing uncertainty in the backscattering coefficient measurementCombined influence of different parameters on the backscattering coeffici
18、ent measurement and especially roughness,biomass and moistureComplexity of the interaction phenomena between the SAR signal and natural surfaces which is not fully understoodGeometry of SAR systemsSAR radiometry correction:Speckle filtering and radiometric enhancementOften a prerequisite before floo
19、d mapping or before classificationSpatial filtering:Mean filterAdaptative filterGamma-MAPLee,frost etc Temporal filtering :-Annual summing-Seasonal summing Optimized temporal filtering:3 local parameters-Textural distance-Radiometric mean-CorrelationSkkSkkraDNraISraI121),(S1),(1),(Temporal filtering
20、 on S images)Optimized temporal filtering =take in account local meanj=1,S.S=number of image in the time seriee,I=Intensity of the pixel(a,r),m=local mean of the neighbourhood of the pixel(a,r),k=weighting coefficient of the image k with:o),(),(.),(1ramramIraIkjkSkkjk.S,/1 Ici.1et k ,0S1kkkkSAR radi
21、ometry correction:Speckle filtering and radiometric enhancementGamma-MAP Filter(widow size=3x3)Simple temporal mean,based on a 7 images time serie Optimized temporal filtering-ISAR radiometry correction:Speckle filtering and radiometric enhancementRadiometry correction:speckle filteringGamma MAP fil
22、terRadiometry correction:speckle filtering by temporal filteringOne ERS imageResult of ERS temporal summing1 SAR Image8 SAR imagesTemporal fusion a tool for classification enhancementNotion of Phase coherenceCoherence value is defined as:The degree of coherence for each pair(s1,s2)of co-registered c
23、omplex values s1,s2 is given by:represents an average,which is estimated by the spatial average over a finite-size windowPhase coherence is closely related to the local deviation of the phase differentialPhase coherence provides information on geophysical change of the observed targetPhase differenc
24、e and coherencePhase difference notion:Two signal of the same wavelength but one is displaced The displacement between corresponding points of the signal are know as phase difference.Phase difference is processed to obtain height and/or motion information o the Earth surface.A source for SEM generat
25、ionssssss121122*Phase coherence sensitivityCoherence by land cover theme:High coherence:bare soil,built-up area,rocksModerate coherence:vegetated areas,cropsLow coherence:forested areasNo coherence:Water bodies,major changeCoherence is related toGeometry of the target Acquisition geometry(notion of
26、baseline)Stability of the target between two successive acquisitionsNotion of successive is broad:can be simultaneous with two decaled antennas,or two days,revisit time,35 days for ERS,etc.SAR coherence data:mutitemporal acquissitionCoherence,Amplitude,Amplitude Changein RGBMultitemporal SAR coheren
27、ce data and Multi-temporal land use classificationMicrowave sensors history1960s First airborne radar systems:SLAR1970s First airborne SAR systems1980s Spaceborne SAR systems:SEASAT,SIR-A,SIR-BDevelopment of Radargrammetry1990s Many radar satellites:JERS-1,ALMAZ for ResearchERS-1,ERS-2(Research/Comm
28、ercial),RADARSAT(Commercial)Highlights:Ocean and Land Surfaces applicationsCommercial applicationsDevelopment of Interferometry2000 ENVISAT(Research/Commercial)Near future:Japanese ALOS,Canadian Radarsat 2/3,Italian Cosmo Skymed,German Terra SAR,Chinese SARERS-1/2w C Band,V V,23o incidencew 100 x 10
29、0 km image with 25m resolutionw Extensive global archive since 91RADARSATw C Band,HH,20o-49o incidencew Range of viewing(20 beam positions)w Range in image coverage(50 x 50 to 500 x 500km)w Range in resolution(8-100m)w Limited archive but good programming serviceENVISAT SARw C Band,HH,VV,HV 20o-49o
30、incidencew Range of viewing(beam positions)w 37 different and mutually exclusive operating modes in high,medium(Wide Swath Mode),w Wide swath coverage;405 km swath with 150 m or 1 km resolution.w Resolution,30 m to 1 kmw Data transfert potential coupling with ArtemisENVISAT SARImage Mode(IM)VV or HH
31、 polarisation images from any of 7 selectable swaths.Swath width between approximately 56 km(swath 7)and 100 km(swath 1)across-track.Spatial resolution of approximately 30 m(for precision product).Alternating Polarisation Mode(AP)Two co-registered images per acquisition,from any of 7 selectable swat
32、hs.HH/VV,HH/HV,or VV/VH polarisation pairs possible.Spatial resolution of approximately 30 m(for precision product).Wide Swath Mode(WS)400 km by 400 km wide swath image.Spatial resolution of approximately 150 m by 150 m for nominal product.VV or HH polarisation.Global Monitoring Mode(GM)Spatial reso
33、lution of approximately 1000 m in azimuth by 1000 m in range for nominal product.Up to a full orbit of coverage.HH or VV polarisation.Wave Mode(WV)A small imagette(dimensions range between 10 km by 5 km to 5km by 5km)is acquired at regular intervals of 100 km along-track.The imagette can be position
34、ed anywhere in an Image Mode swath.Up to two positions in a single swath or in different swaths may be specified,with acquisitions alternating between one and the other(successive imagettes will hence have a separation of 200 km between acquisitions at a given position).HH or VV polarisation may be
35、chosen.Imagettes are converted to wave spectra for ocean monitoring.ENVISAT SARFuture VHR SAR SensorsRADARSAT 2(Near future launch scheduled for February 2002,April 2003 and November 2005)daily revisit VHR SAR mode 3mALOS:L band,Phase array Terra-SAR:constellation X band(potentially a second constel
36、lation L band)Cosmo-Skymed(2004-2006)4 satellites constellation=less than 12 hour revisit timeVHR SAR mode 1-3mChinese X band in 2006SAR VHR =More documented and accurate landscape description =Urban flood mapping =Less confusion between flooded area and neighbourhoodRADARSAT IIStandard mode:25*25mU
37、ltra Fine mode 3*3mTERRA-SAR XLaunch scheduled for March 2005(10-02-04)TERRA-SAR XTERRA-SAR XCOSMO SKYMEDSAR system developed by Italian leadership in conjunction with the optical VHR French Pleiades=Orfeo constellationConstellation of 4 satellite launch schedule 2004-2006 With 4 satellite revisit t
38、ime:12hBand X:Multi polarisation,HH,VV,HV or VHImage modeOne selectable polarizationSPOT_LIGHT:1m or less,10*10 km2H_ IMAGE:few meters,several tens km2WIDE_ REGION(ScanSAR):few tens of meters,hundreds of kmHUGE_REGION(ScanSAR):several tens of meters,hundred km2Two selectable polarizationPING PONG:fe
39、w meters,several tens km2Flood extent mapping based on VHR SAR sensorVHR SAR 50 cmIGN 25 000eRadarsat 12,5 mVHR SAR 0.5 mIGN 25 000eRadarsat 12,5 mFlood extent mapping based on VHR SAR sensorMore documented and accurate landscape descriptionVHR SAR 0.5 mRadarsat 12,5 mFlood extent mapping based on V
40、HR SAR sensorMore documented and accurate landscape descriptionRadar 50 cmIGN 25 000eRadarsat 12,5 mFlood extent mapping based on VHR SAR sensor0.5 m12.5 mVHR SAR,0.5 mIGN 25 000eRadarsat 12,5 mFlood extent mapping based on VHR SAR sensorVery accurate flood extent mappingFundamental for SAR Remote Sensing
