MultidimensionalMultidimensional SAR SAR ImagingImaging::StudiesStudies in the in the FrameworkFramework

of LIMES Projectof LIMES ProjectM. Costantini1, G. Fornaro2, F. Lombardini3, M. Pardini3, F. Serafino2, F. Soldovieri2

1 Telespazio S.p.A.Via Tiburtina, 965 - I-00156 Roma (Italy)

2 IREA-CNRVia Diocleziano,328 I-80124 Napoli (Italy)

3 Dip. di Ingegneria dell’Informazione, Università di PisaVia G. Caruso, 16 - I-56122 Pisa (Italy)

FRINGE 2007 Workshop - European Space Agency

Overview

The LIMES projectMotivation3D tomographic imaging, ERS results4D differential tomography, ERS resultsPerformance evaluation, satellite clustersConclusions

In the framework of LIMES project, SAR 3D/4D Tomography is experimented for supporting Critical Infrastructure SurveillanceCritical regasification plants and pipelines in Spain considered as test areas

Development of satellite-based services providing relevant information and decision-support tools in relation to:

Critical infrastructure monitoringOrganization and distribution of humanitarian relief & reconstructionSurveillance of the EU borders (land and sea)Surveillance and protection of maritime transport for sensitive cargoProtection against emerging security threats

Among the main users involved in the project there are Civil Protection organizations, FRONTEX, EU and International Agencies

LIMESLand and Sea Integrated Monitoring for European Security

Sixth Framework Program GMES Security

Motivation

3D SAR 3D SAR TomographyTomography Multibaseline dataSeparation in elevation of scattering contribution within a single pixelFull 3D Imaging

4D SAR 4D SAR ImagingImaging((differentialdifferential tomographytomography))

Multibaseline multitemporal data Joint elevation-velocity reconstruction

[Pasquali-Prati-Rocca et al., IGARSS ’95] [Reigber-Moreira, IEEE-TGARS ’00]

[Lombardini, IEEE-TGARS ’05]

Surface penetrationSteep ground topography (layover)High spatial density of strong scatterers

Superposition of responses frommultiple scatterers in the same pixel

• accurate determination of scatterer locations• precise tracking of their movement

• SAR interferometry • Multibaseline InSAR• SAR differential interferometry• Multitemporal interferogram stacking• Persistent scatterers

•

Need for more sophisticated techniques

3D Imaging: SAR Tomography

flight d

irectio

n Data acquired at the nth antenna (pass):

SN

Sn

S0

xs

r

z

g0

gn

gN

Need for data calibration (removal of atmospheric variations, scene deformations, …)

Backscatteringprofile

elevation

n-th orthogonalbaseline

After collecting all data, the problem is the inversion of a simple semi-discrete linear operator:

Beamforming SVDAdaptive beam.MUSIC……

The inversion can be affordedwith different algorithms

(linear, regularized, adaptive, parametric, …)

SPATIAL COMPLEX AMPLITUDE SPECTRUMSPATIAL COMPLEX AMPLITUDE SPECTRUM[ Note: equivalent to the classical imaging concept ]

3D: Real data experimentsSan Paolo stadium, Naples

ERS-1/2 data, 63 imagesBaseline span: 1700 m, height resolution 5.5 m Temporal span: ∼ 10 years

range

Singular Value Decomposition(single look)

SVD(5 azimuth looks)

height

azimuth azimuth

Adaptive beamforming(5 azimuth looks)

SAR image

4D: Differential Tomography

flight d

irectio

n

SNB-1,NT-1

Sn,m

S0,0

x s

r

z

rn,m (s)

v

NB : number of tracks per passNT : number of passes

Multistatic system: data acquired at the nth

track and the mth pass:

Elevation-velocitybackscattering

profile

n-th orthogonalbaseline

m-th pass

SNB-1,0

S0,NT-1 g0,0

gn,m

gNB-1,NT-1

gNB-1,0

g0,NT-1

SPATIOSPATIO--TEMPORAL COMPLEX AMPLITUDE SPECTRUMTEMPORAL COMPLEX AMPLITUDE SPECTRUM

After collecting all data, we have again:

2D Beamforming2D SVD2D Adaptive beamforming……

4D: Experimental results (1)

amplitude

Theoretical point-spread function

2D Beamforming

Velocity (mm/yr)

2D Adaptive beamforming

Velocity (mm/yr)

“Mergellina”, NaplesSingle scattering mechanism30 tracksBaseline span: 1066 m, height resolution 8.8 mTime span: ∼ 6 years

-10 -5 0 5 10

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

Doppler (mm/year)

Ele

vatio

n (m

)

2D Capon TOMO-DOPPLER

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

Velocity (mm/yr)

Elev

atio

n(m

)

2D Capon2D Adaptive beamforming

Scatterers well resolved in heightEstimation of deformation velocity consistent with independent measuresReduced SLL w.r.t. 2D Fourier beamforming, but higher sensitivity to miscalibration residualsNo equal velocity constraints (equal velocity case: [Ferretti-Bianchi-Prati-Rocca, EURASIP JASP ’05])

San Paolo stadium, NaplesDouble scattering mechanism

2D Adaptive beam., multilook(9 looks)

-10 -5 0 5 10

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

Doppler (mm/year)

Ele

vatio

n (m

)

2D Beam. TOMO-DOPPLER

0

0.2

0.4

0.6

0.8

1

1.2

Velocity (mm/yr)

Elev

atio

n(m

)

2D Beamforming

-3.1 mm/yr

∼ 10 m

4D: Experimental results (2)

“Vomero”, NaplesERS-1/2, 58 passes, ∼10 years temporal span

- Single scatterers - - Double scatterers -

Scattering mechanisms can be separatedAutomatic single/double scatterer identification also tested

SVD single look

4D: Experimental results (3)

San Paolo Stadium, Naples

- Single scatterers - - Double scatterers -

SVD single look

4D: Experimental results (4)

Imaging Capabilities and Satellite Clusters (1)Typical poor and irregular

baseline/time samplingHigh sidelobes in the 3D/4D High sidelobes in the 3D/4D

reconstructedreconstructed profileprofile

Double speckled compact scatterersSNR = 15, 12 dB32 looks

Elevation ElevationVelocity Velocity

2D Beamforming 2D Adaptivebeamforming

Single track per pass

Elevation ElevationVelocity Velocity

2D Beamforming 2D Adaptivebeamforming

3 tracks per pass

Imaging Capabilities and Satellite Clusters (2)

Acquisition grid (baseline/time) Singular value (SV) distribution

Larger SV dynamic

2 antennas per pass

58 passes, orth. baseline separation 150 m

Accuracy Bounds EvaluationAlgorithm performance judgementCharacterization of precision limits

ToolsTools fromfrom informationinformation theorytheory::

3D Cramér-Rao Lower Bound (CRLB)[Gini-Lombardini-Montanari, IEEE-Tr. on AES ’02]

3D Hybrid CRLB (HCRLB)takes into account possible miscalibration residuals[Pardini-Lombardini-Gini, IEEE-TSP, accepted for publication]

Given a statistical model for the data vector g, bounds can be

evaluated for the 4D estimation of scatterer scatterer elevationselevations and line of line of

sightsight velocitiesvelocitiesERS-1/2, 58 passes10 looksDouble speckled scatterers, large critical baseline(b⊥ TOT /bC = 0.05, classical triangular-shaped spatial decorrelation)Possible temporal decorrelation, τc = 2 months (exponential decorrelation model)[Lombardini-Griffiths, IEE Meeting on RS Sign. Proc. ’98][Rocca-De Zan-Monti Guarnieri-Tebaldini, ENVISAT Symp. ’07]

CRLB Sample Curves

Single track per pass

4D: height

Double scatterer distance in height: 2 resolution unitsRelative motion: 0.7 mm/yr

4D: l.o.s. velocity

[Fully ideal]

-5 0 5 10 15 20 25 3010

-2

10-1

100

101

Signal-to-Noise ratio (dB)

Ver

tical

hei

ght s

tand

ard

devi

atio

n (m

)

4D - CRLB on height estimation - No temporal decorrelation

1 antenna2 antennas

4D: height

NO temporaldecorrelationNO ATMOSPHERE

2 antennas per passBaseline separation: 150 m

Limited advantage in the heightprecision limit, but gain in the SLL

High gain expected in real cases withmiscalibration residuals (atmosphere)HCRLB for 4D: work in progress

ConclusionsIn this work, we have summarize the achievements of 3D/4D SAR imagingwith satellite long-term data.The presented results demonstrate that urban scatterers can be separatedin the elevation/velocity domain by multi-dimensional imaging.By means of numerical tests and analytical bounds we have investigatedthe potentialities of SAR tomography with satellite clusters.

Future systems (e.g. COSMO-Skymed) or cooperative satellite formations (CartWheel, Pendulum, e.g. Tandem-X, ASI Sabrina) are expected in the future

to collect high resolution data with lower temporal separation, or simultaneously.

Thus, the accuracy and performance are expected to increase.