Segmentation-Based Building Analysis from Polarimetric Synthetic Aperture Radar Images [Elektronische Ressource] / Wenju He. Betreuer: Olaf Hellwich
175 Pages
English
Gain access to the library to view online
Learn more

Segmentation-Based Building Analysis from Polarimetric Synthetic Aperture Radar Images [Elektronische Ressource] / Wenju He. Betreuer: Olaf Hellwich

Gain access to the library to view online
Learn more
175 Pages
English

Description

Segmentation-Based Building Analysisfrom Polarimetric Synthetic ApertureRadar Imagesvorgelegt vonMaster of EngineeringWenju Heaus Chinavon der Fakultat IV - Elektrotechnik und Informatikder Technische Universitat Berlinzur Erlangung des akademischen GradesDoktor der Ingenieurwissenschaften- Dr.-Ing. -genehmigte DissertationPromotionsausschuss:Vorsitzender: Prof. Dr. Oliver BrockGutachter: Prof. Dr.-Ing. Olaf Hellwich Prof. Dr.-Ing. Stefan Hinz (KIT)Tag der wissenschaftlichen Aussprache: 15 Juli 2011Berlin 2011D 83AcknowledgementsIn the past four years I have studied Synthetic Aperture Radar (SAR) image processing atthe computer vision and remote sensing group of Technische Universit at Berlin. Professor Dr.-Ing Olaf Hellwich guides me through the study. I sincerely thank him for providing me theopportunity to pursue PhD study in the eld of SAR. He helps me to organize and arrange thestudy. His patience and kindness impress me.I bene t a lot from the discussion meetings in our group. I learn from the way thecolleagues do researches. We share knowledge and have interesting discussions. Constructivesuggestions by colleagues are very helpful. Priv.-Doz. Dr. Andreas Reigber shared his profoundknowledge in SAR data processing to me. He leaded me into this eld by explaining the basicconcepts in details. He guided me to implement the basic and advanced SAR processing tech-niques. Marc J ager guided me into the amazing world of machine learning.

Subjects

Informations

Published by
Published 01 January 2011
Reads 11
Language English
Document size 41 MB

Exrait

Segmentation-Based Building Analysis
from Polarimetric Synthetic Aperture
Radar Images
vorgelegt von
Master of Engineering
Wenju He
aus China
von der Fakultat IV - Elektrotechnik und Informatik
der Technische Universitat Berlin
zur Erlangung des akademischen Grades
Doktor der Ingenieurwissenschaften
- Dr.-Ing. -
genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Prof. Dr. Oliver Brock
Gutachter: Prof. Dr.-Ing. Olaf Hellwich Prof. Dr.-Ing. Stefan Hinz (KIT)
Tag der wissenschaftlichen Aussprache: 15 Juli 2011
Berlin 2011
D 83Acknowledgements
In the past four years I have studied Synthetic Aperture Radar (SAR) image processing at
the computer vision and remote sensing group of Technische Universit at Berlin. Professor Dr.-
Ing Olaf Hellwich guides me through the study. I sincerely thank him for providing me the
opportunity to pursue PhD study in the eld of SAR. He helps me to organize and arrange the
study. His patience and kindness impress me.
I bene t a lot from the discussion meetings in our group. I learn from the way the
colleagues do researches. We share knowledge and have interesting discussions. Constructive
suggestions by colleagues are very helpful. Priv.-Doz. Dr. Andreas Reigber shared his profound
knowledge in SAR data processing to me. He leaded me into this eld by explaining the basic
concepts in details. He guided me to implement the basic and advanced SAR processing tech-
niques. Marc J ager guided me into the amazing world of machine learning. He shared a lot of
interesting papers and new ideas to me. He also helped me in e cient IDL programming and
problems with Linux system.
Dr.-Ing. Hongwei Zheng helped me to settle down during my rst year in Berlin. Lots of
thanks are given to Ronny H ansch. We travelled together to several conferences and had many
interesting discussions. I thank Ms. Marion Dennert for her e orts in arranging my studies and
travels. Dr.-Ing. Volker Rodehorst kindly helped me to revise and print several posters. I would
also like to thank him for solving computer problems for me. Dr. Maxim Neumann answered
me some questions of SAR processing during conferences. Sincere thanks are give to Adam
Stanski, Adhish Prasoon, Andreas Friedrich, Anke Bellmann, Cornelius Wefelscheid, David
Bornemann, Esra Erten, Matthias Heinrichs, Oliver Gloger, Saquib Sarfraz, Stefan Stoinski,
Stephane Guillaso and Ulas Yilmaz.
I am grateful to Technische Universit at Berlin for providing me the scholarship for my
study. I sincerely thank Ms. Roswitha Paul-Walz for arranging my scholarship. I am also grate-
ful to our group and Technische Universit at Berlin for sponsoring me to attend the conferences.
The lectures and posters in the conferences enlarge my view of state-of-the-art developments of
SAR technologies.
Some of my work is inspired by Dr. Derek Hoiem. He also helped me on the implemen-
tation details of the algorithms. The work from Dr. Kevin P. Murphy helped me to implement
Conditional Random Fields and Hidden Markov Model.
Many sincere thanks are given to my friends. The friendship gives me con dence, en-
courage and happiness. Finally, I would like to sincerely thank my parents and brother for
continuous support to my study.
3Abstract
High resolution Synthetic Aperture Radar (SAR) imagery has many applications in urban areas,
e.g. land cover classi cation and building displacement measurement. Buildings are evident in
SAR imagery due to their strong backscattering compared to natural environment. Polarimetric
SAR (PolSAR) imagery combines horizontal and vertical polarizations. It provides polarization
features of scatterers on buildings. Polarimetric decomposition aims to interpret the scattering
process as contributions from several mechanisms, e.g. surface, double bounce and volume
scattering. The scattering characteristics derived from PolSAR imagery can be exploited for
object analysis.
In this thesis we analyze buildings in meter-resolution PolSAR imagery in urban areas.
Segmentation and classi cation of PolSAR imagery are investigated. Polarimetric features,
e.g. amplitude, parameters from polarimetric decomposition and coherence, are extracted. We
adapt state-of-the-art feature extraction, segmentation and classi cation framework for urban
area analysis.
Segmentation provides an initial premise for semantic object analysis. The generated
segments provide spatial support for e cient feature extraction. A good segmentation is critical
for region feature extraction and e cient object detection. We adopt watershed, mean shift,
e cient graph-based segmentation and normalized cuts for PolSAR imagery. These algorithms
produce satisfying segmentation results. The segmentation results are evaluated on ground truth
data.
We propose the integration of probabilistic boundary estimation and segmentation from
PolSAR imagery. One framework is spectral graph segmentation based on probabilistic bound-
ary algorithm. Accurate boundaries are obtained through combining di erent types of gradi-
ents. The segmentation results preserve the weak boundaries. Another framework is occlusion
boundary estimation, in which segmentation and boundary extraction are interleaved. The
segmentation results are of the highest accuracy.
Object extraction is achieved by supervised classi cation of the segments. We extract
polarimetric and e ective low-level features, including texton histogram, histogram of oriented
gradients and sale-invariant feature transform descriptor. Texton histogram is well adapted to
PolSAR imagery. The classi cation aims to group the segments into several semantic classes. We
adopt several strategies for grouping. The rst is Conditional Random Fields, which emphasizes
that neighboring segments are prone to belong to a same class. The second is classi cation based
on multiple segmentations algorithm, which explores the capability of a hierarchy of segmenta-
tions providing spatial support for object evidence extraction. The last strategy is exploiting
building alignment angle and evidence from other objects in a Bayesian detection model. The
appearance of a building in a PolSAR image is in uenced by its alignment angle with respect to
the ight trajectory. We extract e ective features and train classi er to identify building align-
ment angle. Experimental results demonstrate the e ectiveness of these classi cation strategies.
Subaperture analysis is an important tool for SAR data processing. Each subaperture
spans a di erent part of the Doppler spectrum and samples object re ections at di erent az-
imuth look angles. The dependency of object scattering on azimuth look angle is modeled by
Hidden Markov Model (HMM), which describes the behavior variations of buildings across the
subapertures. States in the HMM represent representative centers in the feature space. The
state sequence along the subapertures indicates the scattering dynamics, which is valuable for
the analysis of stationary and non-stationary scatterers. The HMM is also able to classify
buildings from clutter and discriminate between buildings with di erent alignment angles.
4Contents
1. Introduction 9
1.1. Synthetic Aperture Radar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.1.1. Polarimetric SAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.1.2. Interferometric SAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2. Characteristics of SAR Imagery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.1. SAR Data Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2. Advantages of SAR Imagery . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.3. Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3. Buildings in SAR Imagery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3.1. SAR Imagery of Urban Areas . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3.2. Building E ects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.3.3. Building Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.4. Organization and Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2. Polarimetric SAR Imagery 25
2.1. Scattering Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.1. Scattering Coe cient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.2. Scattering Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.3. Second-order Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2. Scattering Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3. Polarimetric Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3.1. Sphere, Diplane and Helix Decomposition . . . . . . . . . . . . . . . . . . 29
2.3.2. Eigenvalue Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.3. Freeman and Durden Decomposition . . . . . . . . . . . . . . . . . . . . . 32
2.4. Polarization Orientation Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.5. Subaperture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.6. Coherence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.7. Polarimetric Speckle Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.8. Polarimetric SAR Classi cation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.8.1. Wishart Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.8.2. Classi cation Based on Scattering Mechanisms . . . . . . . . . . . . . . . 40
2.9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3. Segmentation of PolSAR Imagery 43
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.1.1. Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.1.2. SAR Image Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.1.3. Related Work on SAR Image Segmentation . . . . . . . . . . . . . . . . . 44
3.1.4. Work on PolSAR Image Segmentation . . . . . . . . . . . . . . . 47
3.2. SAR Amplitude Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5Contents
3.2.1. Amplitude Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.2.2. Mixture Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2.3. PolSAR Imagery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.3. Segmentation Algorithms and Experiments . . . . . . . . . . . . . . . . . . . . . 54
3.3.1. Unsupervised Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.3.2. Watershed Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.3.3. Mean Shifttation . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.3.4. E cient Graph-Based Segmentation . . . . . . . . . . . . . . . . . . . . . 62
3.3.5. Normalized Cuts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.4. Multiple Segmentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.5. Evaluation of Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.5.1. Best Spatial Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.5.2. Other Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4. Segmentation based on Probabilistic Boundaries 75
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2. Polarimetric CFAR Edge Detector . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.3. Pb Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.3.1. Gradients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.3.2. Combination of Gradients . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.3.3. Accuracy Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.4. Segmentation based on Probabilistic Boundaries . . . . . . . . . . . . . . . . . . 83
4.4.1. Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.4.2. Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.5. Occlusion Boundary and Segmentation . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5.1. Minimum Merging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5.2. CRF Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5.3. Segmentation from Boundaries . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5.4. Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5.5. Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5. Building Classi cation 97
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2. Related Work in Building Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.1. Coherent Scatterers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.2. Electromagnetic Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.3. Building Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.4. InSAR and PolInSAR Imagery . . . . . . . . . . . . . . . . . . . . . . . . 103
5.3. Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3.1. PolSAR Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3.2. Other Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4. Classi cation using Texton Histogram . . . . . . . . . . . . . . . . . . . . . . . . 111
5.5. using Multiple Segmentations . . . . . . . . . . . . . . . . . . . . . 112
5.6. using Conditional Random Fields . . . . . . . . . . . . . . . . . . . 113
5.6.1. Conditional Random Fields . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.6.2. Classi cation Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6Contents
5.6.3. Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.6.4. Building Layover Detection Results . . . . . . . . . . . . . . . . . . . . . . 117
5.6.5. Shadow Detection Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.7. Analysis of Urban Scene Classi cation . . . . . . . . . . . . . . . . . . . . . . . . 120
5.8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6. Bayesian Building Extraction 123
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
6.2. Building Alignment Classi cation . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.2.1. Building Alignment Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.2.2. Amplitude Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.2.3. Polarimetric Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.2.4. Building Alignment Estimation . . . . . . . . . . . . . . . . . . . . . . . . 127
6.3. Bayesian Model for Building Extraction . . . . . . . . . . . . . . . . . . . . . . . 133
6.3.1. Bayesian Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.3.2. Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.3.3. Inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.3.4. Building Extraction Results . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
7. Urban Area Characterization using Hidden Markov Model 139
7.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
7.2. Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
7.3. Hidden Markov Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
7.4. Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.4.1. Subaperture Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.4.2. Alignment Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.4.3. Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.4.4. HMM Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.4.5. with Log-normal Mixture . . . . . . . . . . . . . . . . . . . . . . . 145
7.5. Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
7.5.1. 8 Subapertures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
7.5.2. 4ertures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
7.5.3. HMM with Log-normal Mixture . . . . . . . . . . . . . . . . . . . . . . . 151
7.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
8. Summary and Future Work 153
8.1. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.2. Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
A. Bibliography 157
B. List of Figures 173
C. List of Tables 175
7Contents
81. Introduction
Synthetic aperture radar (SAR) systems are active sensors that transmit electromagnetic (EM)
microwaves and receive backscattering from ground surfaces. The received signals contain the
information of the scattering process on ground surfaces. Nowadays, advanced spaceborne SAR
systems provide meter-resolution SAR imagery over a huge area, and airborne SAR systems
provide decimeter-resolution imagery. SAR imagery has many advantages due to its peculiar
imaging mechanisms. For instance, active imaging and long wavelength enable the independence
of weather conditions. Polarimetric SAR (PolSAR) and interferometric SAR (InSAR) technolo-
gies substantially improve the capacity of SAR. SAR imagery has great potential in many urban
applications. Building analysis using high resolution SAR imagery has drawn a lot of attention
in recent years.
PolSAR imagery is valuable in interpretation of scattering mechanisms. The scattering of ground
targets depends on polarization. The combination of polarizations helps to identify the distri-
bution of scatterers. Many decomposition and signal processing algorithms, e.g. polarimetric
decomposition, subaperture analysis and coherent scatterer analysis, have been extensively in-
vestigated in the SAR community. The statistical distribution of PolSAR covariance matrix
data is taken into account in many segmentation and classi cation applications.
In this thesis we analyze buildings in meter-resolution PolSAR imagery. Buildings are evident in
SAR data because of their strong backscattering. Building characterization and extraction are
promising and feasible using high resolution airborne SAR imagery. Segmentation methods for
buildings have promising applications in 3D characterization, change detection, reconstruction
and so on. However, inherent characteristics of SAR imaging in urban areas and speckle make
the segmentation problem very di cult. The statistical characteristics of SAR imagery in urban
areas are still in investigation.
Several experiments on urban area characterization using fully polarimetric SAR data have been
conducted in our work. Watershed [145], mean shift [30], e cient graph-based segmentation [62]
and normalized cuts [184] are adapted to PolSAR imagery. Another advanced spectral segmen-
tation [71] is adopted based on probabilistic boundaries, which are extracted by probabilistic
boundary extraction algorithm [141]. Occlusion boundary extraction [99] is able to produce
a segmentation with the highest accuracy. An accurate segmentation allows e ective feature
extraction for object classi cation. Conditional Random Fields [210] and classi cation based
on multiple segmentations [97] are adapted to extract objects in SAR imagery. A Bayesian
framework is proposed by integrating building alignment angle estimation and surface evidence
for better building detection. Furthermore, Hidden Markov Model is adopted to analyze the
variations of building scatterers in subapertures.
91. Introduction
1.1. Synthetic Aperture Radar
A radar sensor transmits electromagnetic waves and receives backscattering signals from any
objects along the propagation path. The interactions of the transmitted signals and the ob-
jects depend on the geometrical con guration, moisture, biomass, surface roughness and other
physical properties. The distance between the sensor and an object can be calculated from
the time delay between transmitting waves and receiving reected waves. Electrically conduc-
tive materials usually scatter strong re ections back to the sensor, and thus become evident in
the observed image. The backscattered signals also provide a rough estimate of object shapes.
Radar is widely used to detect objects in many contexts, e.g. ocean surface wave measurement
and air tra c control.
SAR is di erent from traditional radar systems in that it can provide high spatial resolution
through post-processing after acquiring the raw signals. A very large antenna is needed in
order to obtain a high spatial resolution for radar imagery since a radar adopts lower frequency
microwaves. SAR sensor records re ected signals along its ight trajectory continually. The
received data are focused as if they come from a long antenna. A synthetic aperture is formed in
the focusing process. The synthetic aperture is much longer than the actual aperture. Therefore,
a higher resolution is obtained from a small antenna mounted on aircrafts or satellites. The
typical wavelengths of microwaves used in SAR sensors span from one centimeter through several
meters [58].
A high resolution is crucial for object analysis. More details of geometrical and electromagnetic
features of ground targets are observed in high resolution SAR imagery. Some occlusions are
resolved as well. Spatial averaging is an important step in SAR image preprocessing before it is
ready and reliable for various applications. The averaging improves the stability and accuracy of
SAR signals. However, the smoothing reduces the spatial resolution and may cause severe degra-
dation to object identi cation. For applications in urban areas, spatial resolution plays a critical
role. Informations of geometrical layout and extent of a building are only partially observable
in radar images. Therefore it is very di cult to analyze buildings in SAR imagery. Fortunately,
in decimeter-resolution SAR imagery, many details of building structures are available [146].
1.1.1. Polarimetric SAR
PolSAR and InSAR are two main extensions of SAR technology. Polarization of a wave is the
direction of electric eld which is perpendicular to the direction of wave propagation. SAR
polarimetry exploits the abundant information from multi-polarization acquisitions. A PolSAR
sensor transmits horizontal (H) and vertical (V) polarized signals and receives re ected signals
using both polarizations. Fully polarimetric SAR data consist of four bands, i.e. HH, HV, VH
and VV, where each pair of capital letters represents the polarization modes of sending and
receiving signals. They are formulated in the form of a 2 2 scattering matrix. It is called the
conventional scattering vector. A scattering matrix contains statistical information of scatterer
strength, phase shifts and spatial arrangement of scattering elements within a resolution cell.
For reciprocal targets whose HV and VH components are equal, there are ve degrees of freedom
in the scattering matrix. Scattering mechanism and target characteristics can be retrieved from
the fully polarimetric SAR data. Information extraction from PolSAR data is summarized in
[203].
10