GeomaticsandInformationScienceofWuhanUniversity Vol.40No.11 Nov.2015 DOI: /j.whugis : (2015) SAR 1 1,2

Size: px

Start display at page:

Download "GeomaticsandInformationScienceofWuhanUniversity Vol.40No.11 Nov.2015 DOI: /j.whugis : (2015) SAR 1 1,2"

勘俞
5 years ago
Views:

1 GeomaticsandInformationScienceofWuhanUniversity Vol.40No.11 Nov.015 DOI /j.whugis (015) SAR 极化 SAR 干涉测量 (PolInSAR) 的相干性是反演植被参数的重要信息来源极化空间中不同极化状态对应的相干性分布与复极化相干矩阵的值域相关本文讨论了不同结构的复极化相干矩阵值域的特点分析了不同极化状态下干涉相干性与复极化相干矩阵的值域的关系利用模拟数据和真实全极化 SAR 数据分析了滑动窗口大小和不同散射体对复极化干涉矩阵值域的影响以及复极化干涉矩阵的结构对极化干涉 SAR 相干性分布空间的影响有助于更加准确地获取极化干涉 SAR 最优极化基和估计最优相干性矩阵值域 ; 复相干矩阵 ; 最优相干性 ; 极化干涉 SAR TP751 A SAR (PolInSAR) PolInSAR [1] [3] Po- linsar [4] PolInSAR 1.1 SAR S S = SHH SHV (1) [ S VH S ] VV [5] SAR [1] Pauli K = 1 [(S HH +S VV )(S HH -S VV )S HV ] T (thefieldofvalues) [6-7] 槡 Hilbert [8] Hilbert () Hilbert Po- linsar Hilbert K 1 K [1] Tabb Flynn T 6 [T 6 [59-11] ]=< K1 T 11 K[ ] [K 1 *T K *T Ω1 ] >= [ Ω1 T ] (3) [T 11 ] [T ] [Ω1] [Ω1] 3 3 ;*T SAR 1 [T 11 ]= K 1K *T (01AA11301;011AA10404); (1JJ4035) SAR SAR longjiangping11@163.com

2 40 11 SAR 157 [T ]= K K *T [Ω1]= K 1K *T Ω1 = Ω *T 1 (4) [T 11 ] [T ] ;[Ω1] ω *T [Ω1]ω γ(ω)= ω *T ω =1 槡 ω *T [T 11 ]ωω *T [T ] ω ω1 ω μ1 μ (1) [13-14] [15-17] K 1 K ω1 ω μ1 =ω 1 *T K 1 ; μ =ω *T K (5) 槡 ω *T T 11ωω *T T (5) T 6 *T [ μ[ ] μ 1 μ *T ] = ω 1 *T [T 11 ]ω1 ω 1 *T [Ω1]ω [ ] [T 6 ]= μ1 ω *T [Ω1]ω1 ω *T [T ]ω (4)~ (6) γ [13-18] γ(ω1ω)= γ e i φ = ω *T 1 [Ω1]ω (6) 槡 ω 1 *T[T ]ω1ω *T 11 [T ]ω (7) 1..1 T 6 珟 T 6 珟 T 6 = Π[ I Π I ] =RT6R (8) [ ] R = T-1/ T -1/ ;Π =T -1/ 11 Ω1T -1/ γ =γ(v 1 v )=v *T 1 Πv v *T 1 v =1 (9) v T i= 槡 iωi 槡 T iωi ω i *T 1.3 A n n A. [19] 3 3 F(A)= {x *T Axx C n x *T x =1} (10) Schur [1015] F(A) A U Π =T11-1/ Ω1T -1/ = Hilbert Hilbert λ1 a b 熿燄 (16) F(A) [U ] *T 0 λ c [U ] [0-3] 燀 0 0 λ3 燅 (9) (15) F(Π) [11] F(Π)=γ(v 1 v )=v 1 *T Πv v 1 v C n v 1 *T v 1 =v *T v =1 ( 11) (11) v= T 槡 ω ω H=T -1 T (1) γ =λ(v)=v *T HΠv v *T Hv =1 (13) (13) F(Π)= {γ(v) v C 3 v *T Hv =1}= {v *T HΠv v C 3 v *T Hv =1} (14) A [6] [ ] A = λ1 a 0 λ (15) A Π λ1 λ (a /); λ1=λ A (a /) ; a=0 λ1 λ A λ1 λ ; a=0 λ1=λ A 3 3

3 SAR 1(a) Schur 1(b) SAR SAR 1 SAR (16) Schur 3(a)~3(d) b=0c= λ3 (a) λ3 ; (b) λ3 3(a) ; 3(d) (c) 3 3 (d) 1 Tab.1 TableofSimulatedForestParameters /m 18 3 /( ) 0 /(db m -1 ) /( ) 30 SAR L SIR-C

4 40 11 SAR 159 SAR Pauli 4 6 A B C D E 5 A B A B C D E (d) 4(d) 7(a) 3(d) ; 4(d) 7(b)~ 7(f) 5 6 A B C3 6 A B C3 A C D E 3. 7(b) SAR 7(b) 7(c) 7(d) 7(b) 7(c) φ 0 γ =e i φ 0 (17) (17) 6 D E D E 5 7(e) 7(f) SAR SAR E-SAR Oberpfafen- hofen SAR 7(e) 7(f) 6

1530 015 11 1 Fig.1 TheFieldofValuesofMatrix 4 SIR-C Fig.4 InfluenceofWindowSizestotheFieldof ValuesofCoherenceMatrixwithSIR-CSAR Data 5 Fig.5 TheFieldofValuesofCoherenceMatrixwith Fig.

5 Fig.1 TheFieldofValuesofMatrix 4 SIR-C Fig.4 InfluenceofWindowSizestotheFieldof ValuesofCoherenceMatrixwithSIR-CSAR Data 5 Fig.5 TheFieldofValuesofCoherenceMatrixwith Fig. TheFieldofValuesofCoherenceMatrix SurfaceScater withdiferentformation 6 Oberpfafenhofen Fig.6 ImageoftheOberpfafenhofen 3 SAR Fig.3 InfluenceofWindowSizestotheFieldof ValuesofCoherenceMatrixwithSimulatedSAR Data

6 40 11 SAR 1531 [10]Neumann MReigberAFerro-FamilLetal.Po- 7 linsarcoherencesettheoryandapplication[c]. Fig.7 TheFieldofValuesofCoheretMatrixwith EUSAR006Dresden006 DiferentScatersType [11]Neumann MFerro-FamilLReigberA.Estima- tionofforeststructuregroundandcanopylay- ercharacteristics From Multibaseline Polarimetric InterferometricSAR Data[J].InternationalJour- [1] CloudeSRPapathanassiouKP.PolarimetricSAR Interferometry[J].IEEE Transactionson Geosci- enceand RemoteSensing199836(51) [] LuoHuanminChenErxueChengJianetal.For- estheightestimation Methods UsingPolarimetric SARInterferometry[J].Journalof RemoteSens- ing01014(4)806-81(. SAR [J]. 00-0) 01014(4)806-81) [14] YangZhen.StudiesonSynthetic Aperture Radar [3] Tan Lulu Chen BingYang Ruliang.Improved Interferometric and Polarimetric Interferometric Three-stage Algorithm of Tree Height Retrieval Techniques[D].BeijingChineseAcademyofSci- withpolinsar Data[J].Journalof System Simu- ences003 (. lation 010(4) (. [J]. 010(4) ) [4] WuYirongHong WenWangYanping.TheCur- PolInSAR [15] Neumann MReigber AFerro-Famil L.Data Basedon PolInSAR Coherence Shapes[C].005 IEEEInternationalGeoscienceandRemoteSensing rentstatusandimplicationsofpolarimetricsarin- SymposiumSeoulKorea005 mation Technology0079(5)158-16(. SAR terferometry[j].journalof Electronics & Infor- [J]. 0079(5)158-16) [5] FlynnTTabb MCarandeR.CoherenceRegion ShapeExtractionforVegetationParameterEstima- tioninpolarimetricsarinterferometry[c].00 IEEEInternationalGeoscienceandRemoteSensing SymposiumToronto00 [6] HornR A JohnsonCR.Topicsin MatrixAnaly- sis [M].CambridgeCambridge University Press 1991 [7] Tabb MOrreyJFlynnTetal.PhaseDiversity A DecompositionforVegetationParameterEstima- tion Using Polarimetric SAR Interferometry[C]. ProcEUSARCologneGermany00 [8] HeCWatsonG A.An AlgorithmforComputing thenumericalradius[j].ima Journalof Numer- ical Analysis199717(3)39-34 [9] Tabb MIynn TCarandeR.Estimationand Re- movalofsnr and Scatering Degeneracy Efects fromthepolinsarcoherenceregion[c].the003 IEEEInternationalGeoscienceandRemoteSensing SymposiumToulouseFrance003 nalof RemoteSensing01048(3) [1]SaguesLLopez-SanchezJMFortunyJetal.In- doorexperimentsonpolarimetricsarinterferome- try[j].ieee Transactionson Geoscienceand Re- motesensing00038() [13]ChenBingZhangPing.PolInSAR DataSimulation forinversionstudybasedonscatering Model[J]. Journalof System Simulation0080(8)00-0(. PolInSAR [J]. 0080(8) [D]. 003) [16]ChenErxue.DevelopmentofForestBiomassEsti- mationusingsar Data [J].World Forestry Re-

7 search 19991(6)18-3 (. [J]. [1]DengShaopingLiPingxiangZhangJixianatal (6)18-3) Filtering of Polarimetric SAR Imagery Based on [17]Gong M.CoherenceOptimizationUsingthePolari- Multiplicative Model[J].GeomaticsandInforma- zationstateconformationinpolinsar[j].geoscience and RemoteSensing Leters005(3) [18]BaiLuCaoFangHong Wen.FastApproachto EstimatetheLongestAxisinCoherenceRegionand ItsApplications[J].Journalof Electronics & In- formation Technology0103(3) ( Classification Method Based on Feature Selection. and Adaptive Decision Treefor SAR Image[J]. [J]. 0103(3) ) GeomaticsandInformation Scienceof Wuhan Uni- [19]ChenXiZhangHongWangChao.VegetationPa- versity0137(1)46-49(. rameterextractionusingdualbaselinepolarimetric SAR SARInterferometryData[J].Journalof Electron- [J]. 0137(1)46- ics & Information Technology00830(1) (. [3]ZhangZhongshan Yan QinYuJieetal.Unsu- [J]. pervised Classification of Fuly Polarimetric SAR 00830(1) ) DataBasedonthe PSO Algorithm[J].Geomatics [0]YangJieShiLeiLiPingxiang.Wishart-H/Alpha and Information Science of Wuhan University ClassificationBasedon OptimalCoherenceandPo- larimetricspan [J].Geomaticsand Information Scienceof Wuhan University0137(1) (1)-5) ( ) Wishart-H/Alpha [J]. tion Scienceof Wuhan University011 36(10) (. SAR [J] (10) ) [] He ChuLiu MingXu Lianyu.A Hierarchical 49) 01035(8) (. SAR [J] (8) PolarimetricInterferometrySARCoherenceUsingtheFieldofMatrix LONG Jiangping 1 DING Xiaoli 1 1 DepartmentofSurveyingandGeo-informaticsCentralSouthUniversityChangsha410083China DepartmentofLandSurveyingandGeo-informaticsHongKongPolytechnicUniversityHongKong999077China AbstractPolarimetricinterferometrySAR (PolInSAR)coherenceisimportantinformationwhensol- vingthecoherenceoptimizationproblemsuchaswhenexecutingforestparameterinversion.thedis- tributionofcoherencesthesetofcoherencewithdiferentpolarimetricbasesisrelatedtothefieldof valuesinacomplexcoherencematrix.itisdiscussedthespecialtyofthefieldofvaluesofcomplexco- herencematrixwithdiferentformationandtherelationshipbetweenthecoherencewithdiferentpo- larimetricstatesandthefieldofvaluesofcomplexcoherencematrix.usingsimulatedsardataand realsardatatheinfluenceofthesizesofwindowandscaterstypeonthefieldofvaluesofacom- plexcoherencematrixareevaluated.theresultsindicatethatthefieldofvaluesofcoherencematrix canprovideanew methodtoobtainoptimalpolarimetricbasesandoptimalcoherence. Keywordsthefieldofvalues;coherencematrix;optimizationcoherence;PolInSAR FirstauthorLONGJiangpingPhDspecializesinPolSARandPolInSARdataprocessing. longjiangping11@163.com FoundationsupportTheNationalHi-TechResearchandDevelopmentProgram (863Program)Nos.01AA AA10404; thehunannaturalsciencefoundationprogramno.1jj4035.

07-3.indd

07-3.indd 1 2 3 4 5 6 7 08 11 19 26 31 35 38 47 52 59 64 67 73 10 18 29 76 77 78 79 81 84 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48