| Introduction | p. 1 |
| Background | p. 2 |
| Outline of the Book | p. 3 |
| Stochastic Hyperspectral Measures | p. 3 |
| Subpixel Detection | p. 4 |
| Mixed Pixel classification (MPC) | p. 5 |
| Unconstrained MPC | p. 7 |
| Constrained MPC | p. 7 |
| Automatic Mixed Pixel Classification (AMPC) | p. 8 |
| Hyperspectral Data to be Used in the Book | p. 8 |
| Notations to Be Used in the Book | p. 10 |
| Hyperspectral Measures | p. 13 |
| Hyperspectral Measures for Spectral Characterization | p. 15 |
| Measures of Spectral Variability | p. 15 |
| Spectral Information Measure (SIM) | p. 16 |
| Hidden Markov Model(HMM)-Based Measure | p. 17 |
| Spectral Similarity Measures | p. 20 |
| Commonly Used Measures | p. 20 |
| Distance-Based Measures | p. 20 |
| Orthogonal Projection-Based Measures | p. 20 |
| Spectral Information Divergence (SID) | p. 21 |
| Hidden Markov Model-Based Information Divergence (HMMID) | p. 23 |
| Measures of Spectral Discriminability | p. 23 |
| Relative Spectral Discriminatory ProBability (RSDPB) | p. 24 |
| Relative Spectral Discriminatory PoWer (RSDPW) | p. 24 |
| Relative Spectral Discriminatory Entropy (RSDE) | p. 25 |
| Experiments | p. 26 |
| AVIRIS Data | p. 26 |
| HYDICE Data | p. 31 |
| Conclusions | p. 34 |
| Subpixel Detection | p. 37 |
| Target Abundance-Constrained Subpixel Detection: Partially Constrained Least-Squares Methods | p. 39 |
| Introduction | p. 39 |
| Linear Spectral Mixture Model | p. 40 |
| Orthogonal Subspace Projection (OSP) | p. 41 |
| Sum-to-one Constrained Least Squares Method (SCLS) | p. 44 |
| Nonnegativity Constrained Least Squares Method (NCLS) | p. 45 |
| Hyperspectral Image Experiments | p. 48 |
| Conclusions | p. 50 |
| Target Signature-Constrained Subpixel Detection: Linearly Constrained Minimum Variance (LCMV) | p. 51 |
| Introduction | p. 51 |
| LCMV Target Detector | p. 53 |
| Constrained Energy Minimization (CEM) | p. 54 |
| Target-Constrained Interference-Minimized Filter (TCIMF) | p. 55 |
| Relationship Among Osp, Cem and Tcimf | p. 56 |
| A Comparartive Analysis Between Cem and Tcimf | p. 58 |
| Computer Simulations | p. 58 |
| Hyperspectral Image Experiments | p. 61 |
| Sensitivity of Cem and Tcimf to Level of Target Information | p. 63 |
| Computer Simulations | p. 64 |
| Hyperspectral Image Experiments | p. 67 |
| Real-Time Processing | p. 68 |
| Conclusions | p. 71 |
| Automatic Subpixel Detection: Unsupervised Subpixel Detection | p. 73 |
| Introduction | p. 73 |
| Unsupervised Vector Quantization (UVQ)-Based Algorithm | p. 74 |
| Unsupervised Target Generation Process (UTGP) | p. 75 |
| Unsupervised Ncls (UNCLS) Algorithm | p. 78 |
| Experiments | p. 80 |
| Conclusions | p. 87 |
| Automatic Subpixel Detection: Anomaly Detection | p. 89 |
| Introduction | p. 89 |
| RXD | p. 91 |
| Lptd and Utd | p. 94 |
| Relationship Between Cem and Rxd | p. 97 |
| Real-Time Processing | p. 99 |
| Conclusions | p. 102 |
| Sensitivity of Subpixel Detection | p. 105 |
| Introduction | p. 105 |
| Sensitivity of Target Knowledge | p. 107 |
| Sensitivity of Noise | p. 111 |
| Tscsd | p. 111 |
| Hyperspectral Image Experiments | p. 116 |
| AVIRIS Data | p. 116 |
| HYDICE Data | p. 125 |
| Sensitivity of Anomaly Detection | p. 129 |
| Conclusions | p. 137 |
| Unconstrained Mixed Pixel Classification | p. 139 |
| Unconstrained Mixed Pixel Classification: Least-Squares Subspace Projection | p. 141 |
| Introduction | p. 141 |
| A Posteriori Osp | p. 144 |
| Signature Subspace Projection (SSP) Classifier | p. 144 |
| Target Subspace Projection (TSP) Classifier | p. 146 |
| Oblique Subspace Projection (OBSP) Classifier | p. 147 |
| Unconstrained Maximum Likelihood Estimation Classifier | p. 148 |
| Estimation Error Evaluated by Roc Analysis | p. 150 |
| Signature Subspace Projection (SSP) Classifier | p. 151 |
| Oblique Subspace Projection (OBSP) Classifier | p. 153 |
| Computer Simulations and Hyperspectral Image experiments | p. 153 |
| Computer Simulations | p. 154 |
| Hyperspectral Data | p. 156 |
| Conclusions | p. 159 |
| A Quantitative Analysis of Mixed-to-Pure Pixel Conversion (MPCV) | p. 161 |
| Introduction | p. 162 |
| Conversion of MPC to PPC | p. 162 |
| Mixed-to-Pure Pixel Converter (MPCV) | p. 163 |
| Minimum Distance-Based Classification | p. 164 |
| Fisher's Linear Discriminant Analysis (LDA) | p. 166 |
| Unsupervised Classification | p. 169 |
| Criteria for Target Detection and Classification | p. 169 |
| Comparative Performance Analysis | p. 171 |
| Conclusions | p. 177 |
| Constrained Mixed Pixel Classification | p. 179 |
| Target Abundance-Constrained Mixed Pxiel Classification (TACMPC) | p. 181 |
| Introduction | p. 181 |
| Fully Constrained Least-Squares Approach | p. 183 |
| Fully Constrained Least-Squares Method (FCLS) | p. 183 |
| Unsupervised FCLS Method (UFCLS) | p. 183 |
| Modified Fully Constrained Least-Squares (MFCLS) Approach | p. 184 |
| Computer Simulations and Real Hyperspectral Image Experiments | p. 186 |
| Computer Simulations | p. 186 |
| AVIRIS Image Experiments | p. 188 |
| HYDICE Image Experiments | p. 193 |
| Near Real-Time Implementation | p. 201 |
| Conclusions | p. 205 |
| Target Signature-Constrained Mixed Pixel Classification (TSCMPC): LCMV Classifiers | p. 207 |
| Introduction | p. 207 |
| LCMV Classifer | p. 208 |
| Bowles Et Al.'s Filter Vectors (FV) Algorithm | p. 209 |
| Color Assignment of LCMV Classifiers | p. 211 |
| Extension of Cem Filter to Classifiers | p. 213 |
| Winner-Take-All CEM (WTACEM) Classifier | p. 213 |
| Sum CEM (SCEM) Classifier | p. 213 |
| Multiple-Target CEM (MTCEM) Classifier | p. 213 |
| Target-Constrained Interference-Minimized (TCIM) Classifier | p. 214 |
| Computer Simulations | p. 214 |
| Hyperspectral Image Experiments | p. 218 |
| Real-Time Implementation for LCMV Classifiers | p. 223 |
| Conclusions | p. 227 |
| Target Signature-Constrained Mixed Pixel Classification (TSCMPC): Linearly Constrained Discriminant Analysis (LCDA) | p. 229 |
| Introduction | p. 229 |
| LCDA | p. 230 |
| Whitening Process for LCDA | p. 233 |
| Bowles Et Al.'s Filter Vectors (FV) Algorithm | p. 234 |
| Computer Simulations and Hyperspectral Image Experiments | p. 235 |
| Conclusions | p. 240 |
| Automatic Mixed Pixel Classification (AMPC) | p. 243 |
| Automatic Mixed Pixel Classification (AMPC): Unsupervised Mixed Pixel Classification | p. 245 |
| Introduction | p. 245 |
| Unsupervised MPC | p. 246 |
| Desired Target Detection and Classification | p. 246 |
| Automatic Target Detection and Classification | p. 253 |
| Conclusions | p. 255 |
| Automatic Mixed Pixel Classification (AMPC): Anomaly Classification | p. 257 |
| Introduction | p. 257 |
| Target Discrimination Measures | p. 258 |
| Anomaly Classification | p. 260 |
| Automatic Thresholding Method | p. 260 |
| Analysis on Target Correlation Using Target Discrimination Measures | p. 265 |
| On-Line Implementation | p. 270 |
| Conclusions | p. 274 |
| Automatic Mixed Pixel Classification (AMPC): Linear Spectral Random Mixture Analysis (LSRMA) | p. 277 |
| Introduction | p. 277 |
| Independent Component Analysis (ICA) | p. 279 |
| ICA-Based LSRMA | p. 280 |
| Relative Entropy-Based Measure for ICA | p. 281 |
| Learning Algorithm to Find Separating Matrix W | p. 282 |
| Experiments | p. 284 |
| AVIRIS Image Experiments | p. 284 |
| HYDICE Image Experiments | p. 289 |
| 3-D Roc Analysis for LSRMA | p. 295 |
| Conclusions | p. 302 |
| Automatic Mixed Pixel Classification (AMPC): Projection Pursuit | p. 305 |
| Introduction | p. 305 |
| Projection Pursuit | p. 307 |
| Evolutionary Algorithm (EA) | p. 308 |
| Thresholding of Projection Images Using Zero-Detection | p. 310 |
| Experiments | p. 311 |
| AVIRIS Data Experiments | p. 311 |
| HYDICE Data Experiments | p. 313 |
| Conclusions | p. 318 |
| Estimation for Virtual Dimensionality of Hyperspectral Imagery | p. 319 |
| Introduction | p. 319 |
| Neyman-Pearson Detection Theory-Based Eigen-Thresholding Analysis (Hfc Method) | p. 321 |
| Estimation of Noise Covariance Matrix | p. 323 |
| Residual Analysis (Roger, 1996) | p. 323 |
| Inter/Intra-Band Prediction Noise Estimation: Spatial/Spectral Prediction Noise Estimation (Roger and Arnold, 1996) | p. 325 |
| Noise Estimation-Based Eigen-Thresholding | p. 326 |
| Noise-Whitened HFC (NWHFC) Method | p. 326 |
| Noise Subspace Projection (NSP) | p. 326 |
| AIC and MDL | p. 327 |
| Computer Simulations and Hyperspectral Image Experiments | p. 328 |
| Computer Simulations | p. 328 |
| AVIRIS and HYDICE Image Experiments | p. 330 |
| Conclusions | p. 333 |
| Conclusions and Further Techniques | p. 335 |
| Functional Taxonomy of Techniques | p. 335 |
| Mathematical Taxonomy of Techniques | p. 337 |
| Experiments | p. 339 |
| Roc Analysis for Subpixel Detection and Mixed Pixel Classification | p. 340 |
| Sensitivity Issues | p. 341 |
| Sensitivity to Level of Target Information | p. 341 |
| Sensitivity to Noise | p. 341 |
| Real-Time Implementation | p. 341 |
| Further Techniques | p. 342 |
| Generalized Orthogonal Subspace Projection | p. 342 |
| Convex Cone Analysis | p. 343 |
| Kalman Filter-Based Linear Unmixing | p. 344 |
| Interference-Annihilated Eigen-Analysis | p. 344 |
| Band Selection | p. 345 |
| Linear Mixture Analysis-Based Data Compression | p. 346 |
| Radial Basis Function Neural Network Approach | p. 347 |
| Applications to Magnetic Resonance Imaging | p. 347 |
| Glossary | p. 349 |
| References | p. 353 |
| Index | p. 365 |
| Table of Contents provided by Ingram. All Rights Reserved. |
