SPACEBORNE POLARIMETERIC SYNTHETIC APERTURE RADARdidattica.uniroma2.it/assets/uploads/corsi/145004/8_POLARIMETRIA.pdf · 1 SPACEBORNE POLARIMETERIC SYNTHETIC APERTURE RADAR Università

1

SPACEBORNE POLARIMETERIC SYNTHETIC APERTURE RADAR

Università di Roma Tor VergataDipartimento di Ingegneria Civile e Ingegneria Informatica - DICII

GEOINFORMATION PHD

2

I – Spaceborne Polarimetric Synthetic Aperture Radars

II – Polarization Diversity

III – Polarimetric FeaturesIII – Polarimetric Features

IV - Applications

V - Software

Index

VI - Conclusions

3

SECTION I

Spaceborne Polarimetric Synthetic Aperture Radars

4

Polarimetric SAR systemsPolarimetric SAR systems

Generic polarimetric radar system

5

Multipolarization SystemsMultipolarization Systems- Incoherent systems –

No info about relative phase between orthogonal polarization channels

Can not derive Stokes vector of received field

● Same PRF as Single Pol● Halved azimuth resolution● One Receiving Channel at a time

6

Dual Polarimetric SystemsDual Polarimetric Systems

- Coherent systems –Has info about relative phase between

orthogonal polarization channels

● Same PRF as Single Pol● Same azimuth resolution● Two Receiving Channels

● Doubled PRF with respect to Single Pol

● Reduced range swath (avoid ambiguities)

● One Receiving Channel at a time

7

Full Polarimetric SystemsFull Polarimetric Systems

- Quad Pol systems –Collect all polarimetric information

S matrix

● Doubled PRF with respect to Single Pol● Reduced range swath (avoid ambiguities)● Two Receiving Channels

8

Calibration and CrosstalkCalibration and Crosstalk

[OHH OHV

OVH OVV]=[RHH RHV

RVH RVV ][PHH PHVPVH PVV ][

S HH S HVSVH SVV ][

PHH PHVPVH PVV ][

T HH T HVT VH TVV ]+[

N HH N HV

N VH N VV]

O=R P S PT+NGeneral Model

ScatteringMatrix

Propagationeffects

DistortionEffects on

RX

DistortionEffects on

TX

NoiseObserved

Matrix

9

Calibration and CrosstalkCalibration and Crosstalk

P=[1 00 1]

Propagationeffects

DistortionEffects on RX

DistortionEffects on TX

R=[ 1 δ2δ1 f 1] T=[1 δ4

δ3 f 2]

Channel imbalances

δ1 δ2 δ3 δ4

CrossTalk

On receive On transmit

f 1 f 2Limited by transmitting equal power on both

channels

Special correction techniques required

O=R S T

10

Compact PolarimetryCompact Polarimetry

Reasonable strategy when full polarimetry is precluded

TRANSMIT ONE POLARIZATION RECEIVE TWO ORTHOGONAL POLARIZATIONS

Advantages

Simpler space segment

Lower amount of transmitted data

Can retrieve Stokes vector of the received field

Azimuth resolution not halved

Less risk of crosstalk

11

Compact Polarimetry ModesCompact Polarimetry Modes

45°

H

V

H

V

L L

LR

TX

RX

Circular TransmitLinear Receive

Dual CircularPolarimetricΠ/4

12

Spaceborne POLSAR systemsSpaceborne POLSAR systems

SIR-C - 1994

Present systems

ENVISAT-ASAR | European Space Agency

Alos/PALSAR | Japan Aerospace Exploration Agency

RADARSAT-2 | Canadian Space Agency

COSMO-SkyMed | Italian Space Agency

TerraSAR-X | German Space Agency

May 2012

April 2011

13

PresentPresent

ENVISAT-ASAR – 2002

Nominal Height Revisit Time Inclination

785 km 35 days 98,6°

Image Mode Polarization Mode Swath Spatial Resolution

Imaging, Wave, Alternating Single, AP 100 km 9 – 150 m

Radar Frequency Incidence Angle Polarizations Bandwidth

5,3 Ghz (C-band) 15° - 45° HH,HV,VH,VV N/A

14

PresentPresent

Alos/PALSAR - 2006


691,65 km 46 days 98,16°


Fine, ScanSAR, Polarimetric Single, Dual, Quad 40-70 Km 14-88 m


1,27 Ghz (L-band) 8° - 60° HH,HV,VH,VV 14 – 28 MHz

15

PresentPresent

Radarsat-2 – 2007


546 km 11 days 97°

Radar Frequency Incidence Angle Polarizations PRF

5,32 GHz 20° - 55° HH,HV,VV 3500 Hz


798 km 24 days 98,6°


Fine, Standard, ScanSAR, Quad Single, Dual, Quad 50-500 km 10 – 100 m


5,40 Ghz (C-band) 20° - 40° HH,HV,VH,VV 100 MHz

16

PresentPresent

COSMO-SkyMed - 2007


619 km 12 hours 97,86°


StripMap, ScanSAR, Spotlight Single, PingPong 10 – 200 km 1 – 30 m


9,6 Ghz (X-band) 20° - 59° HH, VV, VH, HV 400 MHz

17

PresentPresent

TerraSAR-X - 2007


514 km 11 days 97,44°


StripMap, ScanSAR, Spotlight Single, Dual, Quad (experimental)

30 - 15 Km 1 - 18 m


9,65 Ghz (X-band) 20° - 45° HH,HV,VH,VV 150 – 300 MHz

18

FutureFuture

Sentinel-1 - 2013

Constellation of two satellite – Revisit Time 1-3 days

Continuity of C-band ESA commitment [ERS-1/2, ENVISAT]

Single or Dual Polarimetric Imaging Modes

Constellation of two satellite – Revisit Time 1-3 days

Continuity of C-band ESA commitment [ERS-1/2, ENVISAT]


Resolutions 5 to 40 m | Swath width 80 to 400 km

19

FutureFuture

Alos-2 - 2013

Improved NESZ, Resolution, Signal-To-Ambiguity Ratio

Follow on of Alos Mission (L-band)

Single, Dual, Full and Compact Polarimetric Modes


20

FutureFuture

Radarsat Constellation – 2014/2015

Constellation of three C-band medium resolution satellites

Follow on of Radarsat-2 mission

Single, Dual, Quad and Compact polarimetric modes


21

FutureFuture

SAOCOM – 2015/2016

Part of SIASGE program with ASI

Two L-band full polarimetric satellites – 8 days repeat cycle

StripMap and TOPSAR and CTLR CP Imaging Modes


22

FutureFuture

COSMO SkyMed Seconda Generazione - 2014/2015

Complete integration with COSMO-SkyMed - CSK

Ensured continuity of X-band ASI commitment


Along Track Interferometry Mode (ATI)

Improved Resolution (500 Mhz Bandwitdh)

Experimental Quad Polarimetric Mode

23

SECTION II

Polarization Diversity

24

E (r , t )=ℜ{E (r )e jω t}

Wave Polarization – The Jones formalismWave Polarization – The Jones formalism

E (r )=E 0 e− j k⋅rPlane wave propagating along z

axis from the source in a lossless homogeneous medium.

E 0 Complex Amplitude k Wave vector

Monocromatic Time-SpaceElectric field

E (r , t )=[E0x cos(ω t−k z+δ x)

E0ycos(ω t−k z+δ y)]=ℜ{[E0x ej δx

E0y ej δy]e− j k z e− jω t}

Jones VectorE

25

Wave Polarization – The Jones formalismWave Polarization – The Jones formalism

Right Circular

VerticalHorizontal

Left Circular

EH=[10] EV=[01]

ER=1

√2 [1− j] EL=

1

√2 [1j]

26

Wave Polarization – The Polarization EllipseWave Polarization – The Polarization Ellipse

E 0xE 0y

=tan (α)

δ=δ x−δy

ψ=(tan (2α)cos (δ)2 )

χ=arcsin (sin (2α)sin (δ)2 )

Rotation angle

Ellipticity

PolarizationEllipse

27

ψ=π

2

χ=0

χ=−(π4 )

Wave Polarization – The Polarization EllipseWave Polarization – The Polarization Ellipse

Right Circular

VerticalHorizontal

Left Circular

ψ=0

χ=0

χ=π

4

28

Wave Polarization – The Stokes formalismWave Polarization – The Stokes formalismExpression of the polarization state using real numbers

F=[I 0S 1S 2S 3]=[

(∣E x∣)2+(∣E y∣)

2

(∣E x∣)2−(∣E y∣)

2

2ℜ{E xE y✳}

2ℑ{E xE y✳ }]=I 0[

1cos(2ψ)cos(2χ)sin (2ψ)cos(2χ)

sin (2χ) ]Stokes Vector

I 02=S 1

2+S 22+S 3

2

m=√S 12+S 22+S 32

I 0

Fully Polarized Wave

Degree of Polarization

I 02>S 1

2+S 22+S 3

2

Partially Polarized Wave

29

Wave Polarization – The Stokes formalismWave Polarization – The Stokes formalism

Poincarè sphere

2ψ 2χLongitude Latitude

Right Circular

VerticalHorizontal

Left Circular

F H=[1100] F V=[

1−100]

F R=[100−1] F L=[

1001]

30

Wave Polarization – The Poincarè SphereWave Polarization – The Poincarè Sphere

L

R

S 1

S 2

S 3 North Pole Left Circular Pol

North Hemisphere Left Elliptical Pol

EquatorLinear Pol

South Hemisphere Right Elliptical Pol

South Pole Right Circular Pol

31

Scattering Operators – Scattering MatrixScattering Operators – Scattering Matrix

S qp=∣S qp∣exp ( j ϕqp)

σ=4π r2(∣E S∣)

2

(∣E I∣)2

σ qp=4π r2 (∣E S

q∣)2

(∣E Ip∣)2

σ0=⟨σ⟩

A0

σ qp=4π(∣S qp∣)2

Radar Cross Section Normalized Radar Cross Section

32


E S=e− jkr

r [S 11 S 12S 21 S 22]E I

E S=e− jkr

rS E I

S ij=∣S ij∣exp ( j ϕij)

Incident Wave Scattered Wave

S=[S 11 S 12S 21 S 22]

2x2 complex Scattering Matrix

E I (r )=E 0I (r)e− j k I⋅r

33


S=1

√4π [√σ11exp ( j ϕ11) √σ12exp ( j ϕ12)√σ21exp ( j ϕ21) √σ22exp ( j ϕ22)]

S=1

√4π [ √σ11 √σ12exp ( j (ϕ12−ϕ11))

√σ12exp ( j (ϕ12−ϕ11)) √σ22exp ( j (ϕ22−ϕ11))]

Scattering Matrix

Removing Absolute Phase (no info) + Considering Reciprocity Theorem

34

Scattering Operators – Mueller MatrixScattering Operators – Mueller MatrixRelation between Incident and Scattered Fields

Scattering Matrix - S

Mueller Matrix - M

Kennaugh Matrix - K

Jones formalism

Stokes formalism

Powers (Mueller using BSA)

S

M K

35

Polarization SynthesisPolarization SynthesisTechnique that permits the calculation of the response of the target for every possible combination of transmitting and receiving antennas.

σ qp=4π r2 (∣E q

r⋅E ps∣)

2

(∣E qr∣)2(∣E p

t∣)2=2π

J qrT K J p

t

(∣E qr∣)2(∣E p

t∣)2=2π jq

rT K j pt

For imaging radars, the individual measurements for each resolution element are statistically related, therefore several power measurements are added to reduce statistical variations.

σ pq0 =2π j q

rT ⟨K ⟩ j pt

36

Polarization ResponsePolarization Response

Large conducting sphere S=[1 00 1]

37


S=[−1 00 1]Dihedral Corner Reflector

38


S=[1 00 1]Trihedral Corner Reflector

39


S=[ cos(α)2 1

2sin (2α)

12sin(2α) sin (α)2 ]

α=0°

Oriented Dipole

40


S=[ cos(α)2 1

2sin (2α)

12sin(2α) sin (α)2 ]

α=45°

Oriented Dipole

41


S=[ cos(α)2 1

2sin (2α)

12sin(2α) sin(α)2 ]

α=90°

Oriented Dipole

42


S=[ 1 − j− j 1 ]Right Helix

43


S=[1 jj −1]Left Helix

44

Scattering Operators – Target VectorsScattering Operators – Target Vectors

S=[SHH SHVSVH SVV ]→ k=

12Tr {ψ S }

ψP={√2[1 00 1], √2[1 0

0 −1], √2[0 11 0]} ψL={2[1 0

0 0], 2√2[0 10 0], 2[0 0

0 1]}

Pauli Spin matrix basis set

Lexicograph matrix basis set

Construction of a vector describing the polarization state of the target

kP=1

√2[SHH+SVV S HH−SVV 2SHV ]

TkL=[S HH √2 S HV SVV ]

T

Physical properties of the target System measurables

45

Scattering Operators – Coherency MatrixScattering Operators – Coherency Matrix

kP=1

√2[SHH+SVV S HH−SVV 2SHV ]

T

T 3=[(∣S HH+SVV∣)

2(SHH+SVV )(S HH−SVV )

✳ (SHH+SVV )(2SHV )✳

(S HH−SVV )(SHH+SVV )✳

(∣S HH+SVV∣)2

(SHH−SVV )(2SHV )✳

(2SHV )(S HH+SVV )✳

(2SHV )(S HH−SVV )✳

(∣2 S HV∣)2 ]

T 3=kP⋅kP✳ T

Coherency matrix

46

Scattering Operators – Covariance MatrixScattering Operators – Covariance Matrix

kL=[S HH √2 S HV SVV ]T

C 3=kL⋅kL✳T

Covariance matrix

C 3=[(∣S HH∣)

2(S HH )(√2 SHV )

✳(SHH )(SVV )

✳

(√2 S HV )(S HH)✳ (∣2SHV∣)

2(√2 SHV )(SVV )

✳

(SVV )(S HH )✳ (SVV )(√2 S HV )

✳(∣SVV∣)

2 ]

47


C 3=12 [1 1 00 0 √21 −1 0 ]T 3[

1 0 11 0 −10 √2 0 ]

Relation between Coherency and Covariance matrices

48


Target Reflection Symmetry Assumption

(S HH)(S HV )✳=(SHV )(SVV )

✳=0

C 3=[(∣S HH∣)

2 0 (SHH )(SVV )✳

0 (∣2SHV∣)2 0

(SVV )(S HH )✳ 0 (∣SVV∣)

2 ]5 unknowns

49


Target Rotation Symmetry Assumption

C 3=[(∣SHH∣)

2 √2 (SHH )(S HV )✳ (∣SHH∣)

2−2 (∣S HV∣)

2

−√2(SHH )(SHV )✳ (∣2SHV∣)

2 √2(S HV )(S HH )✳

(∣SHH∣)2−2 (∣S HV∣)

2−√2 (S HV )(S HH )

✳ (∣SHH∣)2 ]

The Matrix remains unchanged under the transformation

T 3(θ)=R3(θ)T 3 R3(θ)−1

R3(θ)=[1 0 00 cos (2θ) sin (2θ)0 −sin (2θ) cos (2θ)] 3 unknowns

50


Target Azimuthal Symmetry Assumption

C 3=[(∣SHH∣)

2 0 (∣S HH∣)2−2(∣SHV∣)

2

0 (∣2SHV∣)2 0

(∣SHH∣)2−2 (∣S HV∣)

2 0 (∣S HH∣)2 ]

Combines both of the previous properties

2 unknowns

51

SECTION III

Polarimetric Features

52

Polarimetric FeaturesPolarimetric Features

Backscattering Coefficients

Polarimetric Ratios

Copol Phase Difference

Polarimetric Coherence

Eigenvalues Parameters

Target Vector Decomposition Components

Span

53

Features - SpanFeatures - Span

-30 dB >5 dB

SPAN=(∣SHH∣)2+(∣S HV∣)

2+(∣SVH∣)

2+(∣SVV∣)

2

OpticalGoogle Earth ©

RadarSat-2Fine Quad

Total Power

54

Features – Backscattering CoefficientsFeatures – Backscattering Coefficients

-30 dB >5 dB

σHH0

σHV0

σVV0

55

Features – Co-Pol & De-Pol RatiosFeatures – Co-Pol & De-Pol Ratios

-10 dB 8 dB -30 dB 0 dB

σVV0

σHH0

Co-Pol Ratio

σHV0

σHH0 +σVV

0

De-Pol Ratio

56

Features – Co-Pol Phase DifferenceFeatures – Co-Pol Phase Difference

π

π/2π/ 4

0−π/4

−π

−π/2

ϕ=Arg (SHH SVV✶)

57

Features – Polarimetric CoherenceFeatures – Polarimetric Coherence

0 1

ρAA−BB=⟨S AAS BB

✶⟩

√ ⟨∣S AA2∣⟩ ⟨∣S BB2∣⟩

∣ρHH−VV∣ ∣ρHH−HV∣

ππ2

0−π −π2

Arg (ρHH−HV )

58

Features – Target vector ComponentsFeatures – Target vector Components- Imaging Radars -

From Single “Pure” Target to Distributed Target

T=kP⋅k P✳ T

T=1N∑i=1

N

k i⋅k i✳ T

59

Features – Target vector ComponentsFeatures – Target vector Components

Target Decomposition Theorems aim to extract the mean or dominant scattering mechanisms or to separate different scattering contributions.

COHERENT DECOMPOSITIONSMODEL BASED

EIGENVALUES ANALYSIS

ENTROPY/ALPHA/ANISOTROPY

TARGET DICHOTOMY

Pauli, Krogager, Cameron, Touzi Freeman-Durden,Yamaguchi

Cloude

Cloude, Pottier

Huynen, Barnes, Holm

60

Features – Coherent DecompositonsFeatures – Coherent Decompositons

(∣SHH+SVV∣)2[dB ](∣SHH−SVV∣)

2[dB ] (2∣S HV∣)

2[dB ]

RED channel BLUE channel GREEN channel

Pauli

61

Features – Coherent DecompositonsFeatures – Coherent Decompositons

RGB Composite Image

62

Features – Model Based DecompositonsFeatures – Model Based DecompositonsReflection symmetry consideredFreeman-Durden

C 3=C 3S( f s)+C 3D( f d )+⟨C 3V( f v)⟩θ

Canopy scatter from a cloud of randomly oriented dipoles

Double bounce from a pair of orthogonal surfaces

Bragg scattering from a moderately rough surface

PTOT=PS ( f s)+PD( f d )+PV ( f v)

TARGET GENERATORS

63

Features – Model Based DecompositonsFeatures – Model Based Decompositons

PD [dB ] P S [dB ] PV [dB ]


64

Features – Model Based DecompositonsFeatures – Model Based Decompositons

RGB Composite Image

65

Features – Eigenvalues DecompositonsFeatures – Eigenvalues Decompositons

T 3=U 3ΣU 3−1

Hermitian averaged 3x3 Coherency matrix in diagonal form

U 3=[u1 u2 u3 ]

Σ=[λ1 0 00 λ2 00 0 λ3

]λ1≥λ2≥λ3≥0

Eigenvectors of Coherency matrix

Eigenvalues of Coherency matrix

T 3=∑i=1

3

λ iu i⋅u i✳ T=T 01+T 02+T 03

Cloude

66


Single scattering mechanisms“weight” of scattering mechanism

T 01=λ1u1⋅u1✳ T=k 1⋅k1

✳T T 01=[T 1101 T 12

01 T 1301

T 2101 T 22

01 T 2301

T 3101 T 32

01 T 3301]

TARGET GENERATORSExtraction of the dominant

Scattering mechanism

T 3=∑i=1

3

λ iu i⋅u i✳ T=T 01+T 02+T 03

67


T 1101[dB ] T 22

01 [dB ] T 3301[dB ]


68


RGB Composite Image

69

Features – H/alpha/A Features – H/alpha/A ENTROPY/ALPHA/ANISOTROPY

T 3=∑i=1

3

λ iu i⋅u i✳ T

Only one non zero = pure target = single scattering mechanismλ i

All three equal = random target = 3 orthogonal mechanismsλ i

U 3=[u1 u2 u3 ]

Scattering interpreted as a 3 symbol Bernoulli process occurring with pseudoprobabilities

P i=λ i

∑k=1

3

λk ∑ P i=1

with

PROBABILISTIC MODEL FOR SCATTERING

70

Features – H/alpha/A Features – H/alpha/A

U 3=[u1 u2 u3 ]

Parametrized as:

U 3=[cosα1 e

j ϕ1 cosα2 ej ϕ 2 cosα3 e

j ϕ3

sinα1 cosβ1ej(δ1+ϕ1) sinα2cosβ2e

j(δ2+ϕ2) sinα3 cosβ3 ej (δ3+ϕ3)

sinα1 cosβ1ej(γ1+ϕ 1) sinα2 cosβ2 e

j(γ2+ϕ2) sinα3 cosβ3 ej (γ3+ϕ3)]

The probabilistic approach allows the definition of mean scattering mechanism

u=ej ϕ[

cos αsin α cos βe j δ

sin α cos βe j γ] With mean parametersα ,β , δ , γ

71


The only mean parameter following the roll-invariance property

α=∑kPkα k

MEAN ALPHA ANGLE

0⩽α⩽π

2

α≃0

α≃π

4

α≃π

2

Surface scattering

Scattering from acloud of particles

Double-bouncescattering

72


ENTROPY

H=−∑k=1

3

Pk log3 Pk

H≃1

H≃0Pure target

Random target

Degree of mixing of various scattering mechanisms

0⩽H⩽1

73


ANISOTROPY

A=λ 2−λ3λ 2+λ3

In High Entropy Areas, describes the behaviour of the scattering mechanisms apart from the dominant

A≃1

A≃01 high + 2 equal low

2 equal high + 1 low

0⩽A⩽1

74

SECTION IV

Applications

75

ApplicationsApplicationsLand Cover

Surface Parameter Estimation

Agriculture

Forestry

Urban

Ocean

Cryosphere

Target Detection

76

Land CoverLand Cover

Urban Expansion Monitoring

Wildlife Habitat Protection Resource Inventory and Management

Damage Delineation

77

Land CoverLand Cover

Detection of scattering mechanism

Target Decomposition Components & Wishart Classifier for Unsupervised Classification

Geometry recognition allows better classification accuracies

J.-S. Lee, M. Grunes, E. Pottier, and L. Ferro-Famil, “Unsupervised terrain classication preserving polarimetric scattering characteristics," 2004.

F. Xu and Y.-Q. Jin, “Deorientation theory of polarimetric scattering targets and application to terrain surface classication," 2005.

A. Lonnqvist, Y. Rauste, M. Molinier, and T. Hame, “Polarimetric SAR Data in Land Cover Mapping in Boreal Zone," 2010.

Z. Qi, A. G.-O. Yeh, X. Li, and Z. Lin, “A novel algorithm for land use and land cover classication using RADARSAT-2 polarimetric SAR data," 2012.

O. Antropov, Y. Rauste, A. Lonnqvist, and T. Hame, “Polsar mosaic normalization for improved land-cover mapping," 2012.

L. Loosvelt, J. Peters, H. Skriver, B. De Baets, and N. Verhoest, “Impact of reducing polarimetric sar input on the uncertainty of crop classifications based on the random forests algorithm," 2012.

78

Surface Parameter EstimationSurface Parameter Estimation

Important sources for hydrological or meteorological modeling, geology or agriculture monitoring.

Surface Roughness

Soil Moisture

Valuable techniques developed for dual-pol/quad-pol low frequency SAR.

Estimation is not possible using single polarization systems.One of the major issues is how roughness appears to be to radars.

79

Surface Parameter EstimationSurface Parameter Estimation

D. Schuler, J.-S. Lee, D. Kasilingam, and G. Nesti, “Surface roughness and slope measurements using polarimetric sar data," 2002.

I. Hajnsek, E. Pottier, and S. Cloude, “Inversion of surface parameters from polarimetric sar," 2003.

A. Iodice, A. Natale, and D. Riccio, “Retrieval of soil surface parameters via a polarimetric twoscale model," 2011.

N. Baghdadi, R. Cresson, E. Pottier, M. Aubert, M. Zribi, A. Jacome, and S. Benabdallah, “A potential use for the c-band polarimetric sar parameters to characterize the soil surface over bare agriculture fields," 2012.

Y. Kim and J. van Zyl, “A Time-Series Approach to Estimate Soil Moisture Using Polarimetric Radar Data," 2009.

I. Hajnsek, T. Jagdhuber, H. Schon, and K. Papathanassiou, “Potential of Estimating Soil Moisture Under Vegetation Cover by Means of PolSAR," 2009.

H. McNairn, C. Duguay, B. Brisco, and T. Pultz, “The effect of soil and crop residue characteristics on polarimetric radar response," 2002.

X-Bragg model

Two Scale model

Multitemporal Approach

Retrieval using Coherence

80

AgricultureAgriculture

Short-wavelength SAR, such as X- and C-bands, interacts mainly with the top part of canopy layers.

Pedestal height, HV or RR backscattering coefficients.

Characterization of type and amount of cover.

Differential extinction coefficient.

Biophysical parameters retrieval.

Multiple polarizations. Crops classification.

C-band Multipol, C-band Quadpol, X-band Dualpol.

Rice, sugarcane, wheat growth monitoring.

Vegetation Water Content retrieval.

Wetland characterization and flood monitoring.

81

AgricultureAgriculture

J. Lopez-Sanchez, J. Ballester-Berman, and J. Fortuny-Guasch, “Indoor wide-band polarimetric measurements on maize plants: a study of the dierential extinction coecient," 2006.S. Brown, S. Quegan, K. Morrison, J. Bennett, and G. Cookmartin, “High-resolution measurements of scattering in wheat canopies-implications for crop parameter retrieval," 2003.D. Hoekman and M. Vissers, “A new polarimetric classication approach evaluated for agricultural crops," 2003.K. Stankiewicz, “The effciency of crop recognition on ENVISAT ASAR images in two growing seasons," 2006.H. Skriver, F. Mattia, G. Satalino, A. Balenzano, V. Pauwels, N. Verhoest, and M. Davidson, “Crop classication using short-revisit multitemporal sar data," 2011.H. Skriver, “Crop Classication by Multitemporal C- and L-Band Single- and Dual-Polarization and Fully Polarimetric SAR," 2012.H. McNairn, J. Shang, X. Jiao, and C. Champagne, “The contribution of alos palsar multipolarization and polarimetric data to crop classication," 2009.J. Chen, H. Lin, and Z. Pei, ”Application of ENVISAT ASAR Data in Mapping Rice Crop Growth in Southern China," 2007.F. Wu, C. Wang, H. Zhang, B. Zhang, and Y. Tang, “Rice Crop Monitoring in South China With RADARSAT-2 Quad-Polarization SAR Data," 2011.S. Yang, X. Zhao, B. Li, and G. Hua, “Interpreting RADARSAT-2 Quad-Polarization SAR Signatures From Rice Paddy Based on Experiments," 2012.J. Lopez-Sanchez, J. Ballester-Berman, and I. Hajnsek, “First results of rice monitoring practices in spain by means of time series of terrasar-x dual-pol images," 2011.J. Lopez-Sanchez, S. Cloude, and J. Ballester-Berman,”Rice Phenology Monitoring by Means of SAR Polarimetry at X-Band," 2012.W. Koppe, M. L. Gnyp, and C. H “Rice monitoring with multi-temporal and dual-polarimetric TerraSAR-X data," 2012.H. Lin, J. Chen, Z. Pei, S. Zhang, and X. Hu, “Monitoring Sugarcane Growth Using ENVISAT ASAR Data," 2009.G. Satalino, F. Mattia, T. Le Toan, and M. Rinaldi, ”Wheat Crop Mapping by Using ASAR AP Data," 2009.C. Notarnicola and F. Posa, “Inferring Vegetation Water Content From C- and L-Band SAR Images," 2007.B. Marti-Cardona, C. Lopez-Martinez, J. Dolz-Ripolles, and E. Blade-Castellet, ”ASAR polarimetric, multi-incidence angle and multitemporal characterization of Donana wetlands for floode extent monitoring," 2010.R. Touzi, A. Deschamps, and G. Rother, ”Phase of target scattering for wetland characterization using polarimetric c-band sar," 2009.

82

ForestryForestryNeed to develop a biomass cartography of the whole planet. Carbon cycle monitoring.

Frequency dependance of biomass.

P- L- C- band Quadpol (AIRSAR - NASA)

Forest structure

Tree height

Stand age retrieval.

Forestry is an applicative field where lower frequency bands demonstrated to be useful, as the penetration of longer wavelenghts is proved to be essential for biomass retrieval.

Coherence.

Backscattering Coefficients.

Entropy/Alpha/Anisotropy.

Specific field of research and applications: POLinSAR – Polarimetry & Interferometry

83

ForestryForestry

J. Kellndorfer, M. Dobson, J. Vona, and M. Clutter, “Toward precision forestry: plot-level parameter retrieval for slash pine plantations with JPL AIRSAR," 2003.

M.Watanabe, M. Shimada, A. Rosenqvist, T. Tadono, M. Matsuoka, S. Romshoo, K. Ohta, R. Furuta, K. Nakamura, and T. Moriyama, “Forest Structure Dependency of the Relation Between L-Band and Biophysical Parameters," 2006.

H. Wang and K. Ouchi, “Accuracy of the K -Distribution Regression Model for Forest Biomass Estimation by High-Resolution Polarimetric SAR: Comparison of Model Estimation and Field Data," 2008.

D. Hoekman and M. Quinones, “Biophysical forest type characterization in the Colombian Amazonn by airborne polarimetric SAR," 2002.

F. Garestier, P. Dubois-Fernandez, D. Guyon, and T. Le Toan, “Forest Biophysical Parameter Estimation Using L- and P-Band Polarimetric SAR Data," 2009.

S. McNeill and D. Pairman, ”Stand age retrieval in production forest stands in New Zealand using C- and L-band polarimetric Radar," 2005.

Y. Maghsoudi, M. Collins, and D. G. Leckie, “Polarimetric classication of Boreal forest using nonparametric feature selection and multiple classiers," 2012.

S. Maity, C. Patnaik, J. Parihar, S. Panigrahy, and K. Reddy, “Study of physical phenomena of vegetation using polarimetric scattering indices and entropy," 2011.

84

UrbanUrbanCapability of recognizing and separating different scattering mechanisms.

High resolution images

Entropy/Alpha/Anisotropy approach.Polarizaton orientation angle.

Classification of urban/suburban areas.

Circular Pol Correlation Coefficient. Characterization of man-made structures.

Monitoring of damaged areas after earthquakes.

C- and L- band full polarimetric data.

Characterization of different scattering mechanisms. Urban density information extraction.

+

San Francisco X-band QuadPol Stripmap (5m)

85

UrbanUrbanSPAN Circular Correlation

Coefficient

86

UrbanUrban

T. M. Pellizzeri, “Classication of polarimetric SAR images of suburban areas using joint annealed segmentation and H/A/alpha polarimetric decomposition," 2003.

T. Ainsworth, D. Schuler, and J.-S. Lee, “Polarimetric SAR characterization of man-made structures in urban areas using normalized circular-pol correlation coefficients," 2008.

K. Iribe and M. Sato, “Analysis of polarization orientation angle shifts by articial structures," 2007.

M. Watanabe, T. Motohka, Y. Miyagi, C. Yonezawa, and M. Shimada, “Analysis of Urban Areas Affected by the 2011 O the Pacific Coast of Tohoku Earthquake and Tsunami With L-Band SAR Full-Polarimetric Mode," 2012.

X. Li, H. Guo, L. Zhang, X. Chen, and L. Liang, “A New Approach to Collapsed Building Extraction Using RADARSAT-2 Polarimetric SAR Imagery," 2012.

R. Schneider, K. Papathanassiou, I. Hajnsek, and A. Moreira, “Polarimetric and interferometric characterization of coherent scatterers in urban areas," 2006.

M. Fujita and Y. Miho, “Analysis of a microwave backscattering mechanism from a small urban area imaged with sir-c," 2006.

87

OceanOceanShip Tracking

Oil Spill Detection Off-shore Platforms Monitoring

Wind Speed Retrieval

Reflection symmetry properties. Detection of man-made objects.

Entropy/Alpha/Anisotropy – CPD. Oil Spill detection.

88

OceanOcean

M. Migliaccio, A. Gambardella, and M. Tranfaglia, “SAR Polarimetry to Observe Oil Spills," 2007.

M. Migliaccio, A. Gambardella, F. Nunziata, M. Shimada, and O. Isoguchi, “The PALSAR Polarimetric Mode for Sea Oil Slick Observation," 2009.

M. Migliaccio, F. Nunziata, A. Montuori, X. Li, and W. Pichel, “A Multifrequency Polarimetric SAR Processing Chain to Observe Oil Fields in the Gulf of Mexico," 2011.

D. Velotto, M. Migliaccio, F. Nunziata, and S. Lehner, ”Dual-Polarized TerraSAR-X Data for Oil-Spill Observation," 2011.

R. Touzi, R. Raney, and F. Charbonneau, “On the use of permanent symmetric scatterers for ship characterization," 2004.

X. Li and J. Chong, “Processing of Envisat Alternating Polarization Data for Vessel Detection," 2008.

J. Chen, Y. Chen, and J. Yang, “Ship detection using polarization cross-entropy," 2009.

B. Zhang, W. Perrie, P. W. Vachon, X. Li, W. G. Pichel, J. Guo, and Y. He, “Ocean vector winds retrieval from c-band fully polarimetric sar measurements," 2012.

89

CryosphereCryosphere

Snow Characterization Glaciers Monitoring Ship routing

VV-HH backscattering L- X-band. Ice thickness retrieval.

Cross polarization channel. Depol. Discrimination between FYI/ MYI

C-band Full pol. Classification of glaciers.

90

CryosphereCryosphere

K. Nakamura, H.Wakabayashi, K. Naoki, F. Nishio, T. Moriyama, and S. Uratsuka, “Observation of sea-ice thickness in the sea of Okhotsk by using dual-frequency and fully polarimetric airborne SAR (pi-SAR) data," 2005.

K. Partington, J. Flach, D. Barber, D. Isleifson, P. Meadows, and P. Verlaan, “Dual-Polarization C-Band Radar Observations of Sea Ice in the Amundsen Gulf," 2010.

J.-W. Kim, D. jin Kim, and B. J. Hwang, “Characterization of arctic sea ice thickness using high resolution spaceborne polarimetric sar data," 2012.

L. Huang, Z. Li, B.-S. Tian, Q. Chen, J.-L. Liu, and R. Zhang, “Classication and snow line detection for glacial areas using the polarimetric SAR image," 2011.

J. Sharma, I. Hajnsek, K. Papathanassiou, and A. Moreira, “Polarimetric decomposition over glacier ice using long-wavelength airborne polsar," 2011.

M. Trudel, R. Magagi, and H. Granberg, “Application of Target Decomposition Theorems Over Snow-Covered Forested Areas," 2009.

91

Target DetectionTarget Detection

Target Ship Moving Object

Change inside the sceneUse of polarimetric properties or configurations for detection.

Reflection Symmetry. Man-made targets detection.

Phase difference between crosspol. Ambiguities detection.

Sensitivity of polarimetric coherence. Recognition similar features in different images.

Land cover classification.

Change detection.Polarization state conformation.

92

Target DetectionTarget Detection

N. Wang, G. Shi, L. Liu, L. Zhao, and G. Kuang, “Polarimetric sar target detection using the reflection symmetry," 2012.

C. Liu and C. Gierull, “A new application for polsar imagery in the field of moving target indication/ship detection," 2007.

A. Marino, S. Cloude, and I. Woodhouse, “A polarimetric target detector using the Huynen fork," 2010.

A. Marino, S. Cloude, and I. Woodhouse, “Detecting Depolarized Targets Using a New Geometrical Perturbation Filter," 2012.

A. Marino, S. R. Cloude, and J. M. Lopez-Sanchez, “A New Polarimetric Change Detector in Radar Imagery," 2012.

M. Qong, “Polarization state conformation and its application to change detection in polarimetric sar data," 2004.

93

Application MaturityApplication Maturity

DOMAIN APPLICATION/PRODUCT MATURITY NOTE

FORESTRY

ABOVE GROUND BIOMASS MEDIUM POLINSAR

STAND HEIGHT HIGH POLINSAR

VERTICAL STRUCTURE MEDIUM POLINSAR

THEMATIC MAPS HIGH

CHANGE DETECTION HIGH

AGRICULTURE

CROP TYPE MAPPING MEDIUM

SOIL MOISTURE HIGH

PHENOLOGY DETERMINATION MEDIUM

FLOODING MAPPING MEDIUM

CRYOSPHERE

SNOW VOLUME MEDIUM

LAND ICE EXTINCTION LOW

SEA ICE SURFACE CHARACTERIZATION LOW

URBANMAPPING/CLASSIFICATION MEDIUM

SUBSIDENCE MEDIUM

OCEANOIL SLICK DETECTION MEDIUM

METALLIC TARGETS MEDIUM

Results of ESA POLinSAR2013 – February 2013

94

SECTION V

Software

95

ESA PolSAR ProESA PolSAR ProLicence Free

Type Standalone application

Sensors Every polarimetric airborne/spaceborne sensor

Analysis Almost every polarimetric technique is implemented

Website http://earth.eo.esa.int/polsarpro/

96

ASF MapReadyASF MapReadyLicence Free

Type Standalone application

Sensors ALOS-Palsar Radarsat-2 TerraSAR-X CEOS-1 format

Analysis Pauli Compositon Sinclair Compositon

Cloude-Pottier Classification (8 and 16 classes)

Sinclair Compositon

Entropy/Alpha/Anisotropy Wishart Classifier

Freman/Durden

Faraday Rotarion Compensation

Website http://www.asf.alaska.edu/downloads/software_tools

97

Exelis SARScapeExelis SARScapeLicence Commercial

Type ENVI module

Sensors

Analysis Polarimetric Calibration

Polarization Signature

Entropy/Alpha/Anisotropy Decomposition and Classification

Pauli Composition

Polarization Synthesis

Most of commercial spaceborne sensors supported

Website http://www.exelisvis.com/ProductsServices/ENVI/SARscape.aspx

98

Radar ToolsRadar ToolsLicence Free

Type IDL Library

Sensors

Analysis


Website http://radartools.berlios.de/

Point Target Analysis

Speckle Filtering


Decomposition Analysis

Edge Detectors Polarimetric Descriptors

Basis Transform Subaperture Analysis

99

Racurs Photomod RadarRacurs Photomod RadarLicence Commercial

Type Photomod Module

Sensors

Analysis Polarimetric Descriptors

Sinclair Composition


Pauli Composition

Classical Decompositions

Website http://www.racurs.ru/?page=86

ALOS-Palsar Radarsat-2 TSX CEOS-1 format CSK

100

PCI GeomaticaPCI GeomaticaLicence Commercial

Type PCI Module

Sensors

Website

Scattering, Covariance, Coherency, Kennaugh Matrices

Correlation Coeff.


Pauli Composition

Pedestal Height


Total Power and Ratios

Phase Differences

Classical Decompositions (Freeman, VanZyl, Huynen, ...)

Speckle Filters

http://www.pcigeomatics.com/products/geomatica-2013

101

SECTION VI

Conclusions

102

Conclusions – Scattering and PolarizationConclusions – Scattering and PolarizationEM Scattering – Wavelength dependance

Surface Roughness

Volume Penetration

Smooth Rough Very RoughIncident

Coherent

Incoherent

L band C band X band

103

Conclusions – Scattering and PolarizationConclusions – Scattering and Polarization

L

C

X

San Francisco Airport Multifrequency effect of Model-Based Freeman Decomposition

104

L

C

X

Treasure IslandYerba Buena Island

Multifrequency effect of Model-Based Freeman Decomposition


105


L

C

X

Golden Gate Bridge Multifrequency effect of Model-Based Freeman Decomposition

106

Conclusions – ModelsConclusions – Models

k σ z

k l c

EM Surface Scattering ModelsSmall PerturbationModel

Integral EquationMethod

Geometric Optics Approximation

σ z - Rms surface height variationsl c - Surface autocorrelation length

k - Wavenumber

107

Full vs Dual PolFull vs Dual Pol

Full Polarimetry Dual Polarimetry

Target ScatteringCharacterization

Range Ambiguities Doubled Data Tx/Rx

Doubled Azimuth Resolution

Double Swath Width

Complete Polarimetric Info

Halved Swath Width

HH-VV Phase HV/VH Information

Incomplete Polarimetric Info

No HH-VV Phase and HV/VH

- Geometry - Material

- Depolarization

- Cross-Polarization

Partial Target ScatteringCharacterization

- Material

108

Conclusions – Other NSAConclusions – Other NSA

Sentinel-1A/B

Radarsat Constellation Mission

Alos-2

TerraSAR-L / TanDEM-L

SAOCOM

C-band Dual-Pol

C-band Dual/Compact/Quad-Pol

L-band Dual/Compact/Quad-Pol

L-band TanDEM

L-band Dual/Compact/Quad-Pol

“We have arrived at the door-steps of the golden age of polarimetric radar imaging.”

- Lee, Pottier - Polarimetric Radar Imaging, CRC 2009

SPACEBORNE POLARIMETERIC SYNTHETIC APERTURE RADARdidattica.uniroma2.it/assets/uploads/corsi/145004/8_POLARIMETRIA.pdf · 1 SPACEBORNE POLARIMETERIC SYNTHETIC APERTURE RADAR Università

Documents