Optical Remote Sensing Advances in Signal Processing and Exploitation Techniques

✍ Scribed by Prasad, Saurabh(Editor);Bruce, Lori M;Chanussot, Jocelyn

• Publisher: Springer Berlin Heidelberg
• Year: 2011
• Tongue: English
• Leaves: 344
• Series: Augmented vision and reality 3
• Category: Library

✦ Synopsis

Optical remote sensing relies on exploiting multispectral and hyper spectral imagery possessing high spatial and spectral resolutions respectively. These modalities, although useful for most remote sensing tasks, often present challenges that must be addressed for their effective exploitation. This book presents current state-of-the-art algorithms that address the following key challenges encountered in representation and analysis of such optical remotely sensed data. Challenges in pre-processing images, storing and representing high dimensional data, fusing different sensor modalities, pattern classification and target recognition, visualization of high dimensional imagery.

✦ Table of Contents

2…Fusion of Optical and Radar Data......Page 2
3.1 The Aim......Page 4
3.1 Experimental Setup......Page 6
Cover......Page 1
Contents......Page 7
Preface......Page 5
3.1.3 Experiment 3: Effects of Misregistration at Different Scales......Page 8
1 Introduction......Page 9
1…Optical Remote Sensing: The Processing Chain......Page 10
4…Conclusions......Page 11
4.2 Linear and Non-Linear Sample Similarity Measures......Page 13
3.2 Neural Network-Based Spectral Unmixing......Page 14
References......Page 15
1…Introduction......Page 17
5.1 Registration Noise Identification......Page 19
Optical Remote Sensing......Page 3
2…Optical Remote Sensing: Key Challenges for Signal Processing and Effective Exploitation......Page 12
3.3 Automatic Selection and Labeling of Training Samples......Page 16
2.2 Lossy, Lossless, Near-Lossless......Page 20
2.4 Image Distribution......Page 21
6.1 Data Set Description......Page 22
2.1 Data Acquisition Process......Page 18
2.5 Data Availability......Page 23
3.1 Prediction-Based......Page 24
3.2 Vector Quantization......Page 25
3.3.1 Transform......Page 26
3.3.2 Coding......Page 27
5.3 Experimental Results......Page 28
3.4 Lossy to Lossless......Page 29
4.1 Why Bothering with Lossy Compression?......Page 30
4.2 Quality Evaluation......Page 31
4.3 Making Comparison Easier......Page 33
6…Conclusion......Page 35
Abstract......Page 38
1…Introduction......Page 39
2…Compressive-Projection Principal Component Analysis (CPPCA)......Page 40
2.1 Overview of CPPCA......Page 41
2.2 The CPPCA Algorithm......Page 42
2.2.1 Eigenvector Recovery......Page 43
2.2.2 Coefficient Recovery......Page 44
3…Compressed Sensing (CS)......Page 45
4…Empirical Comparisons on Hyperspectral Imagery......Page 46
4.1 Performance of Single-Task and Multi-Task CS......Page 47
4.2 Performance of CPPCA and CS......Page 49
4.3 Execution Times......Page 52
5…Conclusions......Page 53
Abstract......Page 56
1…Introduction......Page 57
2.1 Mathematical Representation......Page 59
2.2 Reduced Resolution Imaging......Page 62
3.1 Improving SNR Using Hadamard Multiplexing......Page 66
3.2 Variable Resolution Hyperspectral Sensing......Page 68
4…Summary......Page 70
References......Page 71
1…Introduction......Page 72
2.1 Information Loss......Page 73
2.2 Metrics......Page 74
2.4 Color Saturation and Neutrals......Page 76
2.5 Color Blindness......Page 77
3…Some Solutions......Page 78
3.1 Optimized Basis Functions......Page 79
3.2 Adapting Basis Functions......Page 81
3.3 White Balance......Page 84
4…Conclusions and Open Questions......Page 85
References......Page 86
Abstract......Page 87
2…Image Construction......Page 89
3…Comparative Visualization Techniques......Page 90
3.2 Soft Classification Visualization......Page 91
3.3 Double Layer Visualization......Page 92
4…Experimental Design and Settings......Page 93
5.1.1 Perceptual Edge Detection......Page 95
Results......Page 96
Task......Page 97
Result......Page 98
Task......Page 99
Results......Page 101
6…Discussion and Conclusions......Page 102
References......Page 103
Abstract......Page 105
1…Introduction......Page 106
2…The Proposed Framework......Page 108
2.1 Subspace Identification: Partitioning the Hyperspectral Space......Page 109
2.2.1 Linear Discriminant Analysis (LDA)......Page 112
2.2.2 Kernel Discriminant Analysis (KDA)......Page 113
2.3 Classifier......Page 115
2.4 Decision Fusion......Page 116
3.1 Handheld Hyperspectral Data......Page 117
3.2 Airborne Hyperspectral Data......Page 118
4…Experimental Setup and Results......Page 119
4.1.1 Experiment 1: MCDF with LDA Based Pre-processing at the Subspace Level......Page 120
4.1.2 Experiment 2: MCDF with KDA Based Pre-processing at the Subspace Level......Page 121
4.2 Experiments with Aerial HSI Data......Page 124
5…Conclusions, Caveats and Future Work......Page 125
References......Page 126
Abstract......Page 129
1…Introduction......Page 130
2.1 Fundamental Properties......Page 131
2.2 Opening and Closing by Reconstruction......Page 132
2.3 Attribute Filters......Page 134
3.1 Morphological Profiles......Page 137
3.2 Attribute Profiles......Page 139
3.3 Experimental Results and Discussion......Page 140
4.1 Problem of Extending the Morphological Operators to Multi-tone Images......Page 143
4.2 Extended Morphological Profile......Page 144
4.3 Extended Attribute Profiles......Page 145
4.4 Experimental Results and Discussion......Page 146
5…Conclusion......Page 149
Acknowledgments......Page 150
Abstract......Page 153
1…Introduction......Page 154
2.1 Image Data......Page 155
2.3 Training Data......Page 156
3.1 Decision Fusion Using Hierarchical Tree Structure......Page 158
3.2 Decision Fusion Using the Hierarchical Tree and Class Membership Values......Page 160
3.3 Class-Dependent Neural Networks Ensemble......Page 161
4…Accuracy Assessment......Page 162
4.1 Comparison of Classification Results......Page 163
5.1 Results of Various Tested Classifiers......Page 164
5.2 Results of Class Dependent Neural Networks......Page 169
5.3 Results of Decision Fusion Using Hierarchical Tree Structure......Page 170
5.4 Results of Hierarchical Tree Coupled with Probability Labels......Page 172
5.5 The Assessment of Significance of the Accuracy Values......Page 173
6…Conclusions......Page 174
References......Page 175
Abstract......Page 177
1.1 Classification with Kernels......Page 178
1.3 Feature Extraction with Kernels......Page 179
2.1 Measuring Similarity with Kernels......Page 180
2.3 Basic Operations with Kernels......Page 181
2.5 Kernel Development......Page 182
3.1 Support Vector Machine......Page 183
3.2 nu -Support Vector Machine......Page 184
3.3 Support Vector Data Description......Page 186
3.5 Kernel Fisher’s Discriminant......Page 187
3.6 Experimental Results for Supervised Classification......Page 188
3.6.2 {\varvec \nu} -SVM versus OC-SVM......Page 189
3.6.3 Support Vector versus Fisher’s Discriminant......Page 190
3.7.2 Semisupervised Regularization Framework......Page 191
3.7.3 Laplacian Support Vector Machine......Page 192
3.7.4 Transductive SVM......Page 193
4…Kernel Methods in Biophysical Parameter Estimation......Page 194
4.1 Support Vector Regression......Page 195
4.2 Relevance Vector Machines......Page 196
4.3 Gaussian Processes......Page 198
4.4 Experimental Results......Page 199
5…Kernel Methods for Feature Extraction......Page 200
5.1.1 Principal Component Analysis......Page 201
5.1.2 Partial Least Squares......Page 202
5.2.1 Kernel Principal Component Analysis......Page 203
5.2.2 Kernel Partial Least Squares......Page 204
6.1 Multiple Kernel Learning......Page 205
6.3 Structured Learning......Page 206
References......Page 207
Abstract......Page 213
1…Introduction......Page 214
2…Nonlinear Manifold Learning for Dimensionality Reduction......Page 215
2.1 Dimensionality Reduction Within a Graph Embedding Framework......Page 216
2.2.2 Kernel Principal Component Analysis (KPCA)......Page 217
2.3.1 Locally Linear Embedding (LLE)......Page 218
2.3.2 Local Tangent Space Alignment (LTSA)......Page 219
2.4 Supervised Local Manifold Learning......Page 220
3.1 Botswana Hyperion Data (BOT)......Page 221
3.2 Kennedy Space Center AVIRIS Data (KSC)......Page 222
3.4 ACRE ProspectTIR Data (ACRE)......Page 223
4.1 Performance of Dimensionality Reduction Methods (DR) for BOT Hyperion Data......Page 224
4.2 Comparison of DR Methods for BOT, KSC, IND PINE, and ACRE Sites......Page 228
4.3 Manifold Coordinates for DR Methods......Page 232
5…Summary and Conclusions......Page 235
References......Page 238
Abstract......Page 241
1…Introduction......Page 242
2.1 Problem Formulation......Page 244
2.2 Endmember Extraction......Page 245
2.2.1 N-FINDR......Page 246
2.2.2 Orthogonal Subspace Projection (OSP)......Page 247
2.2.4 Automatic Morphological Endmember Extraction (AMEE)......Page 248
2.2.5 Spatial Spectral Endmember Extraction (SSEE)......Page 249
2.2.6 Spatial Pre-Processing (SPP)......Page 251
2.3 Unconstrained Versus Constrained Linear Spectral Unmixing......Page 252
3.1 Problem Formulation......Page 253
3.2 Neural Network-Based Spectral Unmixing......Page 254
3.3 Automatic Selection and Labeling of Training Samples......Page 256
4.1 First Experiment: AVIRIS Hyperspectral Data......Page 257
4.2 Second Experiment: DAIS 7915 and ROSIS Hyperspectral Data......Page 260
4.2.1 Data Description......Page 261
4.2.2 Fractional Abundance Estimation Results......Page 263
5…Parallel Implementation Case Study......Page 265
6…Conclusions and Future Research......Page 269
References......Page 270
Abstract......Page 274
1…Introduction......Page 275
2…Notation and Background......Page 276
3…Analysis of Registration Noise Properties......Page 278
3.1 Experimental Setup......Page 279
3.1.1 Experiment 1: Effects of Increasing Misregistration on Unchanged Pixels......Page 280
3.1.3 Experiment 3: Effects of Misregistration at Different Scales......Page 281
3.2 Properties of RN in VHR Images......Page 283
4…Proposed Technique for the Adaptive Estimation of the Registration Noise Distribution......Page 288
5…Proposed Change-Detection Technique Robust to Registration Noise......Page 291
5.1 Registration Noise Identification......Page 292
5.2 Context-Sensitive Decision Strategy for the Generation of the Final Change-Detection Map......Page 294
6.1 Data Set Description......Page 295
6.2 Estimation Results......Page 297
6.3 Change-Detection Results......Page 299
7…Discussion and Conclusion......Page 302
References......Page 303
1…Introduction......Page 305
2.1 Component Substitution Methods......Page 307
2.2 Multiresolution Methods......Page 308
2.3 Selected Methods for Testing on HS+Pan Images......Page 309
2.4 Evaluation of Spatial Enhancement Methods......Page 311
4.1.1 GIHS......Page 314
4.1.3 HPF-P......Page 315
4.1.4 GMMSE......Page 317
4.3 PCA......Page 319
4.4 Kernel PCA......Page 322
4.5 Linearity Preserving Projection (LPP)......Page 325
5…Conclusions......Page 329
Abstract......Page 332
2…Fusion of Optical and Radar Data......Page 333
3.1 The Aim......Page 335
3.2 Remote Sensing as a Tool......Page 336
3.3 Decision-Level Fusion......Page 341
4…Conclusions......Page 342
References......Page 343