Multidimensional SAR SAR Imaging Imaging: : Multidimensional - - PowerPoint PPT Presentation

Multidimensional Multidimensional SAR SAR Imaging Imaging: : Studies Studies in the in the Framework Framework

• f LIMES Project
• f LIMES Project
• M. 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-CNR

Via Diocleziano,328 I-80124 Napoli (Italy)

3 Dip. di Ingegneria dell’Informazione, Università di Pisa

Via G. Caruso, 16 - I-56122 Pisa (Italy)

FRINGE 2007 Workshop - European Space Agency

Overview

The LIMES project Motivation 3D tomographic imaging, ERS results 4D differential tomography, ERS results Performance evaluation, satellite clusters Conclusions

In the framework of LIMES project, SAR 3D/4D Tomography is experimented for supporting Critical Infrastructure Surveillance Critical 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 monitoring Organization and distribution of humanitarian relief & reconstruction Surveillance of the EU borders (land and sea) Surveillance and protection of maritime transport for sensitive cargo Protection against emerging security threats Among the main users involved in the project there are Civil Protection

• rganizations, FRONTEX, EU and International Agencies

LIMES

Land and Sea Integrated Monitoring for European Security

Sixth Framework Program GMES Security

Motivation

3D SAR 3D SAR Tomography Tomography

Multibaseline data Separation in elevation of scattering contribution within a single pixel Full 3D Imaging

4D SAR 4D SAR Imaging Imaging ( (differential differential tomography tomography) )

Multibaseline multitemporal data Joint elevation-velocity reconstruction

[Pasquali-Prati-Rocca et al., IGARSS ’95] [Reigber-Moreira, IEEE-TGARS ’00] [Lombardini, IEEE-TGARS ’05]

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

Superposition of responses from multiple 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 direction

Data acquired at the nth antenna (pass):

SN Sn S0 x s r z g0 gn gN Need for data calibration (removal of atmospheric variations, scene deformations, …)

Backscattering profile elevation n-th orthogonal baseline

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

SVD Adaptive beam. MUSIC …… The inversion can be afforded with different algorithms (linear, regularized, adaptive, parametric, …) SPATIAL COMPLEX AMPLITUDE SPECTRUM SPATIAL COMPLEX AMPLITUDE SPECTRUM

[ Note: equivalent to the classical imaging concept ]

3D: Real data experiments

San Paolo stadium, Naples

ERS-1/2 data, 63 images Baseline 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 direction SNB-1,NT-1 Sn,m S0,0 x s r z rn,m (s) v

NB : number of tracks per pass NT : number of passes

Multistatic system: data acquired at the nth track and the mth pass:

Elevation-velocity backscattering profile n-th orthogonal baseline m-th pass

SNB-1,0 S0,NT-1 g0,0 gn,m gNB-1,NT-1 gNB-1,0 g0,NT-1

SPATIO SPATIO-

• TEMPORAL COMPLEX AMPLITUDE SPECTRUM

TEMPORAL COMPLEX AMPLITUDE SPECTRUM

After collecting all data, we have again:

2D Beamforming 2D SVD 2D Adaptive beamforming ……

4D: Experimental results (1)

amplitude

Theoretical point-spread function

2D Beamforming Velocity (mm/yr) 2D Adaptive beamforming Velocity (mm/yr)

“Mergellina”, Naples Single scattering mechanism

30 tracks Baseline span: 1066 m, height resolution 8.8 m Time span: ∼ 6 years

• 10
• 5

5 10

• 50
• 40
• 30
• 20
• 10

10 20 30 40 50 60 Doppler (mm/year) Elevation (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) Elevation (m) 2D Capon 2D Adaptive beamforming

Scatterers well resolved in height Estimation of deformation velocity consistent with independent measures Reduced SLL w.r.t. 2D Fourier beamforming, but higher sensitivity to miscalibration residuals No equal velocity constraints (equal velocity case: [Ferretti-Bianchi-Prati-Rocca, EURASIP JASP ’05])

San Paolo stadium, Naples Double scattering mechanism

2D Adaptive beam., multilook (9 looks)

• 10
• 5

5 10

• 50
• 40
• 30
• 20
• 10

10 20 30 40 50 60 Doppler (mm/year) Elevation (m) 2D Beam. TOMO-DOPPLER 0.2 0.4 0.6 0.8 1 1.2

Velocity (mm/yr) Elevation (m) 2D Beamforming

• 3.1 mm/yr

∼ 10 m

4D: Experimental results (2)

“Vomero”, Naples

ERS-1/2, 58 passes, ∼10 years temporal span

• Single scatterers -
• Double scatterers -

Scattering mechanisms can be separated Automatic 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 sampling High sidelobes in the 3D/4D High sidelobes in the 3D/4D reconstructed reconstructed profile profile

Double speckled compact scatterers SNR = 15, 12 dB 32 looks

Elevation Elevation Velocity Velocity 2D Beamforming 2D Adaptive beamforming

Single track per pass

Elevation Elevation Velocity Velocity 2D Beamforming 2D Adaptive beamforming

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 Evaluation

Algorithm performance judgement Characterization of precision limits Tools Tools from from information information theory theory: : 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 elevations elevations and line of line of sight sight velocities velocities

ERS-1/2, 58 passes 10 looks Double 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 units Relative motion: 0.7 mm/yr

4D: l.o.s. velocity

[Fully ideal]

• 5

5 10 15 20 25 30 10

• 2

10

• 1

10 10

Signal-to-Noise ratio (dB) Vertical height standard deviation (m) 4D - CRLB on height estimation - No temporal decorrelation 1 antenna 2 antennas

4D: height

NO temporal decorrelation NO ATMOSPHERE

2 antennas per pass

Baseline separation: 150 m Limited advantage in the height precision limit, but gain in the SLL High gain expected in real cases with miscalibration residuals (atmosphere)

HCRLB for 4D: work in progress

Conclusions

• In this work, we have summarize the achievements of 3D/4D SAR imaging

with satellite long-term data.

• The presented results demonstrate that urban scatterers can be separated

in the elevation/velocity domain by multi-dimensional imaging.

• By means of numerical tests and analytical bounds we have investigated

the 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,

• r simultaneously.

Thus, the accuracy and performance are expected to increase.