This course gives keys to understand the SAR image and specificities: geometry, speckle, penetration capabilities, layovers, multipath, dielectric properties. Advanced modes: polarimetry, interferomety and POLINSAR are also presented.

1. WHY IS A SAR IMAGE
« DIFFERENT »?
La nature crée
des différences,
la société en fait
des inégalités.
Tahar Ben Jelloun

2.  Advanced modes of SAR images
 Radiometry
 Interferometry
 Polarimetry
 POLINSAR
 Important features
 Geometry
 Speckle
 Electrical properties
2
OUTLINE

3. 3
Important features of the wave
Carrier frequency
Propagation direction (incidence)
polarization
)(cm
1.0 1 10 100
)(GHzf
300 30 3 3.0
Ku Ka X L PSC
hˆ
vˆ
kˆ
hˆ
nˆ
vˆ
hˆ
kˆ
nˆ
hˆHorizontal polarization
(RADARSAT)
Vertical polarization
(ERS)

4. 4
A CLASSIC SAR IMAGE
one antenna
Measurement:
One complex
Coefficientantenna Image 1
i
j
1
S
Aim : 2 D imaging
Only absolute value
is used

5.  The basic measurement made by a SAR is a complex number
S (amplitude and phase).
SLC Single Look Complex
 Main observable:
A is the amplitude image, I=A2 is the intensity image
the phase of a single image is not meaningful)
 The radar Cross Section is defined as:
𝜎 = 4𝜋 𝑅2
𝑃𝑠
𝑃𝑖
R is the radar-target distance
Pi is the incident power, Ps is the power scattered by the target 5
THE SAR IMAGE

6. The image is seen as a
picture.
Pixels are numbers
Image is affected by
speckle noise
Most commonly used:
intensity image
6
WHAT IS A SAR IMAGE ?

7. QUESTION
Low signal : black
or white ?

8. 8
RADIOMETRIC IMAGE
diffusion
Bright points
Specular
Bright areas are produce by strong radar response and darker areas are from
weak radar responses.

9. 9
ASAR (ENVISAT) IMAGE
What are the white
points on the sea ?

10. 10
Quality image parameters
- Resolution
- Pixel sampling

11. Associate images A, B
and C to following
processing:
1. Constrast
enhancement
2. High pass filtering
3. Segmentation
enhancement
IMAGE PROCESSING
A B
C D

12. 0
1
2
3
4
5
6
7
8
x 10
4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.5
1
1.5
2
x 10
5
0 50 100 150 200 250
0
0.5
1
1.5
2
2.5
3
3.5
4
x 10
5
0 50 100 150 200 250
0
1
2
3
4
5
6
7
8
x 10
5
0 50 100 150 200 250
FIND THE TRUE HISTOGRAM
Landes, RAMSES (ONERA)

13. 0
1
2
3
4
5
6
7
8
x 10
5
0 50 100 150 200 250
0
0.5
1
1.5
2
2.5
3
3.5
4
x 10
4
0 50 100 150 200 250
0
0.5
1
1.5
2
2.5
3
3.5
x 10
5
0 50 100 150 200 250
0
0.5
1
1.5
2
2.5
3
3.5
4
x 10
5
0 50 100 150 200 250
FIND THE TRUE HISTOGRAM
Toulouse, SETHI (ONERA)

14. 14
INTERFEROMETRIC SAR (INSAR)
Two antennas
Measurement:
two complex
coefficients
h
Antenna 2
Antenna 1
Image 1
i
j
Image 2
i
j
2
S
1
S
Aim : 3 D cartography

15. H
h
iobs
iinc



2
2
2
1
*
21
2,1



2
1


2 complex signals
Interferometric coherence
(Schwarz inequality)10 2,1  

φ
||
■ h is given by φ=arg()
■ coherence level: |  |

18. 18
1 polarimetric antenna
1
Image 1
i
j








 11
11
1
vvvh
hvhh
SS
SS
S
Mesure:
One 2x2 complex matrix
Polarimetric SAR (POLSAR)

19. 19
1
0
0°180°
0°180°
What does the radar “see”?
 VVHH


23. THE GENERAL BENEFIT OF
POLARIMETRY
Polarimetry helps
classification, but are we
able to understand
polarimetric behaviour ?

24. BASIC POLARISATION
24
The linear basis : H and V polarizations













1
0
0
1
VH

25. 25
The circular basis : L and R polarizations













1
1 j
L
j
R
BASIC POLARISATION

26. CIRCULAR POLARIZATION:
 Circular polarization : H and V with pi/2 phase
26

27. ANY POLARIZATION
 Elliptic
polarization
 Jones vector (a,b)
27

28. 28
How targets can modify polarization
General properties of a medium :
• Change in intensity (absorption)
• Change in phases (refraction)
absorption refraction

29. 29
How targets can modify polarization
When absorption and refraction depend on orientation…
Linear diattenuation
Linear retardater

30. SINCLAIR MATRIX
 complex-valued 2x2 matrix that transforms the polarization
of a plane EM wave incident upon a target to the polarization
of the wave scattered from the target
30







vV
hH
S
S
S
vH
hV
S
S







0
0
1
1
S







10
01
S






10
01
S
EH
EH
EH
EV
EV
EV

31. THE SCATTERING VECTOR
 Pauli basis
31











0
0
1
k











0
1
0
k











1
0
0
k
2S
2
1
hV












 vVhH
vVhH
SS
SS
k

32. QUIZZ
 Color representation
32











0
0
1
k











0
1
0
k











1
0
0
k

33. TWO TYPES OF POLARIMETRIC
BEHAVIOR
 Deterministic
 Non deterministic
33
Saturation level
What are the gray
(non deterministic)
areas ?

34. 34
2 polarimetric antennas
h
Image 1
i
j








 11
11
1
vvvh
hvhh
SS
SS
S
Image 2
i
j








 22
22
2
vvvh
hvhh
SS
SS
S
Mesurement:
two 2x2 complex
matrices
Aim: information about
rhe vertical structure
Interferometric and polarimetric SAR (POLINSAR)

36. Unlike aerial photographs and satellite images which are
passive remote sensing systems
in active systems such as radar, the brightness or darkness
of the image is dependent on the portion of the transmitted
energy that is returned back to the radar from targets on the
surface
TARGET INTERACTION AND IMAGE SIGNATURES

37. SPECKLE

38. RADAR SPECKLE
All radar images appear with some degree of what
we call radar speckle. Speckle 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 resolution cell.
Speckle reduction can be achieved in two ways:
-multi-look processing
-spatial filtering.
Degrade resolution :
must be balanced with the amount
of detail required.

39. 39
THE PHYSICAL ORIGIN OF SPECKLE
Resolution cells are made up of many scatterers with different phases, leading
to interference and the noise-like effect known as speckle.

40. Probability density distribution of speckle:
Intensity image: exponential distribution
Amplitude image: Rayleigh distribution
40
STATISTICS OF SPECKLE

41.  Radar images are formed coherently and therefore inevitably
have a “noise-like” appearance
 Implies that a single pixel is not representative of the
backscattering
“Averaging” needs to be done
Averaging means we also get a decrease in spatial resolution by
the same factor (N)
41
SPECKLE SUMMARY

42. GEOMETRIC EFFECTS
Shadow
Layover/foreshortening
Multipath
Echo superpositions

43.  Layover : B' is seen before A'
 Foreshortening : C'D' is a very short distance
43
LAYOVER AND FORESHORTENING

44. 44
www.intro2radar.html
Thees animarions were created by I. Woodhouse.

45. SHADOW IS MORE OF A
PROBLEM AT FAR RANGE
SHADOW

46.  Viaduc de Millau
THE DARK SNAKE…

47.  Both scatterers are seen in the same pixel.
47
ECHO SUPERPOSITION

48.  Double bounce (left) and triple bounce (right)
48
MULTIPATH

49. 49
URBAN AREAS
Tokyo River Island
Which is the range
axis ?
The azimut axis ?
How would you
evaluate the height
of the buildings ?

50.  Will the radar see the car ?
50
HIDDEN TARGET

51. 51
Targets may appear at strange positions…

52. 52
WHAT IS THIS TRIPLE LINE?

53. PROPERTIES OF THE
TARGET

54. 54
The response to radar energy by the target is primarily
dependent on three factors:
Radar viewing and surface geometry relationship
Surface roughness of the target
Moisture content and electrical properties of the
target

55. 

cos8
h
BRIGHTNESS = ROUGHNESS
Intermediate


cos25
h


cos4,4
h
Smooth
Rough
θi
θr
Smooth surface
Moderately rough
surface
Very rough surface
A surface is classified as smooth or
rough by comparing its surface height
deviation with wavelength.

56. 56
1
0
Gray levels for different roughness

57.  What is dark area in this image, why is it dark?
57
INTERPRETATION OF BRIGHTNESS

58. 58
PENETRATION PROPERTIES
f
P
L

C
0
90

Penetration ability: depends on wavelength

59. 59
X-BAND AND L-BAND
Which image corresponds to the shorter wavelength ?
The longer ?

60. Which color
corresponds to
the shorter
wavelength ?
The longer ?
COLOR CODE WITH FREQUENCY

61. Different vegetation types (e.g., desert, grasslands, forests or frozen tundra)
will all have different backscatter properties.
dielectric constant = the amount of water that the soil contains.
Dry soil = low dielectric constant = little radar energy Saturated soil =
strong reflector.
Moist and partially frozen soils =intermediate values.
DIELECTRIC PROPERTIES

62. Material
Dielectric
constant
Vacuum 1 (by definition)
Air 1.00054
Paper 3.5
Pyrex glass 4.7
Water (20°) 80.4
Most common materials have dielectric constants 1-100
Affecting the absorption and propagation of electromagnetic waves .
Dielectric constant controlled by the amount of moisture content
Increasing the moisture content reduces
the penetration of the radar signal beneath
the soil and vegetation canopy.
DIELECTRIC PROPERTIES

63. GET USED TO OTHER
IMAGES

64.  Brétigny sur Orge (91 Essonne)
OPTICS VERSUS RADAR (B&W)
• Which one is the optical, which one is the radar image?

65.  Toulouse
OPTICS VERSUS RADAR (B&W)

66. 66
WHAT IS COLOR REPRENSENTATIVE FOR ?
TERRASAR-X, Baltimore

67. 67
FALSE COLOR COMPOSITION
Color codes:
1- Polarization
2- interferometry
(elevation)
3- Classification results
based on distribution
evaluation
4- Frequency
TERRASAR-X, Dubai

68. 68
COLOR CODE ?
San Francisco Bay