Image analysis, classification and change detection in remote sensing : with algorithms for Python

Cover......Page 1
Half Title......Page 2
Title Page......Page 3
Copyright Page......Page 4
Table of Contents......Page 5
Preface to the First Edition......Page 12
Preface to the Second Edition......Page 14
Preface to the Third Edition......Page 16
Preface to the Fourth Edition......Page 18
Author Biography......Page 20
1: Images, Arrays, and Matrices......Page 22
 1.1 Multispectral satellite images......Page 25
 1.2 Synthetic aperture radar images......Page 28
 1.3 Algebra of vectors and matrices......Page 31
  1.3.1 Elementary properties......Page 32
  1.3.2 Square matrices......Page 34
  1.3.4 Symmetric, positive definite matrices......Page 36
  1.3.5 Linear dependence and vector spaces......Page 37
 1.4 Eigenvalues and eigenvectors......Page 38
 1.5 Singular value decomposition......Page 40
 1.6 Finding minima and maxima......Page 42
 1.7 Exercises......Page 48
 2.1 Random variables......Page 52
  2.1.1 Discrete random variables......Page 53
  2.1.2 Continuous random variables......Page 54
  2.1.3 Random vectors......Page 57
  2.1.4 The normal distribution......Page 60
  2.1.5 The gamma distribution and its derivatives......Page 62
  2.2.1 Random samples......Page 65
  2.2.2 Sample distributions and interval estimators......Page 68
 2.3 Multivariate distributions......Page 71
  2.3.1 Vector sample functions and the data matrix......Page 72
  2.3.2 Provisional means......Page 74
  2.3.3 Real and complex multivariate sample distributions......Page 76
 2.4 Bayes’ Theorem, likelihood and classification......Page 78
 2.5 Hypothesis testing......Page 81
  2.6.1 One independent variable......Page 86
  2.6.2 Coefficient of determination (R2)......Page 88
  2.6.3 More than one independent variable......Page 89
  2.6.4 Regularization, duality and the Gram matrix......Page 93
 2.7 Entropy and information......Page 94
  2.7.1 Kullback–Leibler divergence......Page 96
  2.7.2 Mutual information......Page 97
 2.8 Exercises......Page 98
 3.1 The discrete Fourier transform......Page 104
 3.2 The discrete wavelet transform......Page 109
  3.2.1 Haar wavelets......Page 110
  3.2.2 Image compression......Page 114
  3.2.3 Multiresolution analysis......Page 117
 3.3 Principal components......Page 124
  3.3.1 Principal components on the GEE......Page 126
  3.3.2 Image compression and reconstruction......Page 128
  3.3.4 Dual solution......Page 132
 3.4 Minimum noise fraction......Page 133
  3.4.1 Additive noise......Page 134
  3.4.2 Minimum noise fraction via PCA......Page 137
  3.5.1 Maximum autocorrelation factor......Page 138
  3.5.2 Noise estimation......Page 140
 3.6 Exercises......Page 144
 4.1 The Convolution Theorem......Page 148
 4.2 Linear filters......Page 153
 4.3 Wavelets and filter banks......Page 156
  4.3.1 One-dimensional arrays......Page 157
  4.3.2 Two-dimensional arrays......Page 162
  4.4.1 Valid kernels......Page 165
  4.4.2 Kernel PCA......Page 170
 4.5 Gibbs–Markov random fields......Page 173
 4.6 Exercises......Page 177
 5.1 Lookup tables and histogram functions......Page 180
  5.2.1 Sobel filter......Page 182
  5.2.2 Laplacian-of-Gaussian filter......Page 185
  5.2.3 OpenCV and GEE algorithms......Page 187
  5.2.4 Invariant moments......Page 192
 5.3 Panchromatic sharpening......Page 198
  5.3.1 HSV fusion......Page 199
  5.3.3 PCA fusion......Page 200
  5.3.4 DWT fusion......Page 201
  5.3.5 À trous fusion......Page 202
  5.3.6 A quality index......Page 205
  5.4.1 Speckle statistics......Page 206
  5.4.2 Multi-look data......Page 209
  5.4.3 Speckle filtering......Page 214
 5.5 Topographic correction......Page 221
  5.5.1 Rotation, scaling and translation......Page 222
  5.5.2 Imaging transformations......Page 223
  5.5.3 Camera models and RFM approximations......Page 224
  5.5.4 Stereo imaging and digital elevation models......Page 226
  5.5.5 Slope and aspect......Page 231
  5.5.6 Illumination correction......Page 232
 5.6 Image–image registration......Page 237
  5.6.1 Frequency domain registration......Page 238
  5.6.2 Feature matching......Page 240
  5.6.3 Re-sampling with ground control points......Page 244
 5.7 Exercises......Page 246
6: Supervised Classification Part 1......Page 252
 6.1 Maximizing the a posteriori probability......Page 254
 6.2 Training data and separability......Page 255
 6.3 Maximum likelihood classification......Page 260
  6.3.1 Naive Bayes on the GEE......Page 261
  6.3.2 Python scripts for supervised classification......Page 262
 6.4 Gaussian kernel classification......Page 266
 6.5 Neural networks......Page 269
  6.5.1 The neural network classifier......Page 274
  6.5.2 Cost functions......Page 277
  6.5.3 Backpropagation......Page 279
  6.5.4 A deep learning network......Page 285
  6.5.5 Overfitting and generalization......Page 289
  6.6.1 Linearly separable classes......Page 291
  6.6.2 Overlapping classes......Page 297
  6.6.3 Solution with sequential minimal optimization......Page 299
  6.6.4 Multiclass SVMs......Page 300
  6.6.5 Kernel substitution......Page 301
 6.7 Exercises......Page 305
 7.1 Postprocessing......Page 310
  7.1.2 Probabilistic label relaxation......Page 311
  7.2.1 Accuracy assessment......Page 314
  7.2.2 Accuracy assessment on the GEE......Page 319
  7.2.3 Cross-validation on parallel architectures......Page 320
  7.2.4 Model comparison......Page 323
 7.3 Adaptive boosting......Page 327
 7.4 Classification of polarimetric SAR imagery......Page 333
  7.5.1 Spectral mixture modeling......Page 335
  7.5.2 Unconstrained linear unmixing......Page 338
  7.5.3 Intrinsic end-members and pixel purity......Page 339
  7.5.4 Anomaly detection: The RX algorithm......Page 340
  7.5.5 Anomaly detection: The kernel RX algorithm......Page 343
 7.6 Exercises......Page 347
8: Unsupervised Classification......Page 350
 8.1 Simple cost functions......Page 351
 8.2 Algorithms that minimize the simple cost functions......Page 353
  8.2.1 K-means clustering......Page 354
  8.2.2 Kernel K-means clustering......Page 359
  8.2.3 Extended K-means clustering......Page 362
  8.2.4 Agglomerative hierarchical clustering......Page 365
  8.2.5 Fuzzy K-means clustering......Page 368
 8.3 Gaussian mixture clustering......Page 370
  8.3.1 Expectation maximization......Page 371
  8.3.3 Partition density......Page 374
  8.4.1 Multiresolution clustering......Page 375
  8.4.2 Spatial clustering......Page 378
 8.5 A benchmark......Page 381
 8.6 The Kohonen self-organizing map......Page 383
 8.7 Image segmentation and the mean shift......Page 387
 8.8 Exercises......Page 389
9: Change Detection......Page 396
 9.1 Naive methods......Page 397
 9.2 Principal components analysis (PCA)......Page 399
  9.2.1 Iterated PCA......Page 401
  9.2.2 Kernel PCA......Page 403
 9.3 Multivariate alteration detection (MAD)......Page 405
  9.3.1 Canonical correlation analysis (CCA)......Page 406
  9.3.2 Orthogonality properties......Page 409
  9.3.3 Iteratively re-weighted MAD......Page 410
  9.3.4 Scale invariance......Page 412
  9.3.5 Correlation with the original observations......Page 413
  9.3.6 Regularization......Page 415
  9.3.7 Postprocessing......Page 417
 9.4 Unsupervised change classification......Page 418
 9.5 iMAD on the Google Earth Engine......Page 420
 9.6 Change detection with polarimetric SAR imagery......Page 422
  9.6.1 Scalar imagery: the gamma distribution......Page 423
  9.6.2 Polarimetric imagery: the complex Wishart distribution......Page 426
  9.6.3 Python software......Page 430
  9.6.4 SAR change detection on the Google Earth Engine......Page 434
 9.7 Radiometric normalization of visual/infrared images......Page 436
  9.7.1 Scatterplot matching......Page 437
  9.7.2 Automatic radiometric normalization......Page 440
 9.9 Exercises......Page 443
 A.1 Cholesky decomposition......Page 448
 A.2 Vector and inner product spaces......Page 450
 A.3 Complex numbers, vectors and matrices......Page 451
  A.4.1 Recursive linear regression......Page 453
  A.4.2 Orthogonal linear regression......Page 455
 A.5 Proof of Theorem 7.1......Page 458
 B.1 The Hessian matrix......Page 462
  B.1.1 The R-operator......Page 463
  B.1.2 Calculating the Hessian......Page 466
  B.2.1 Conjugate directions......Page 468
  B.2.2 Minimizing a quadratic function......Page 470
  B.2.3 The algorithm......Page 472
 B.3 Extended Kalman filter training......Page 476
  B.3.1 Linearization......Page 477
  B.3.2 The algorithm......Page 478
 C.1 Installation......Page 484
  C.2.2 earthengine......Page 485
 C.4 Python scripts......Page 486
  C.4.2 atwt.py......Page 487
  C.4.4 classify.py......Page 488
  C.4.5 crossvalidate.py......Page 489
  C.4.7 dispms.py......Page 490
  C.4.9 eeMad.py......Page 491
  C.4.11 eeWishart.py......Page 492
  C.4.13 em.py......Page 493
  C.4.15 gamma_filter.py......Page 494
  C.4.17 iMad.py......Page 495
  C.4.19 kkmeans.py......Page 496
  C.4.21 kpca.py......Page 497
  C.4.24 meanshift.py......Page 498
  C.4.27 pca.py......Page 499
  C.4.29 radcal.py......Page 500
  C.4.31 registerms.py......Page 501
  C.4.32 registersar.py......Page 502
  C.4.34 sar_seq.py......Page 503
  C.4.36 som.py......Page 504
  C.5.2 omnibus_run......Page 505
  C.5.6 utilities......Page 506
Mathematical Notation......Page 508
References......Page 510
Index......Page 522