Image Processing and GIS for Remote Sensing: Techniques and Applications

Following the successful publication of the 1st edition in 2009, the 2nd edition maintains its aim to provide an application–driven package of essential techniques in image processing and GIS, together with case studies for demonstration and guidance in remote sensing applications. The book therefore has a 3 in 1 structure which pinpoints the intersection between these three individual disciplines and successfully draws them together in a balanced and comprehensive manner.

The book conveys in–depth knowledge of image processing and GIS techniques in an accessible and comprehensive manner, with clear explanations and conceptual illustrations used throughout to enhance student learning. The understanding of key concepts is always emphasised with minimal assumption of prior mathematical experience.

The book is heavily based on the authors own research. Many of the author–designed image processing techniques are popular around the world. For instance, the SFIM technique has long been adopted by ASTRIUM for mass–production of their standard Pan–sharpen imagery data. The new edition also includes a completely new chapter on subpixel technology and new case studies, based on their recent research.

Essential Image Processing and GIS for Remote Sensing 11

Overview of the Book 11

Part I. Image Processing 13

Chapter 1 Digital image and display 14

1.1 What is a digital image? 14

1.2 Digital image display 15

1.2.1 Monochromatic display 15

1.2.2 Tristimulus colour theory and RGB (red, green, blue) colour display 16

1.2.3 Pseudo colour display 18

1.3 Some key points 20

1.4 Questions 20

Chapter 2 Point operations (Contrast enhancement) 21

2.1 Histogram modification and look up table (LUT) 21

2.2 Linear contrast enhancement (LCE) 24

2.3 Logarithmic and Exponential contrast enhancement 26

2.31 Logarithmic contrast enhancement 26

2.32 Exponential contrast enhancement 27

2.4 Histogram equalisation (HE) 28

2.5 Histogram matching (HM) and Gaussian stretch 29

2.6 Balance contrast enhancement technique (BCET) 30

2.7 Clipping in contrast enhancement 33

2.8 Tips for interactive contrast enhancement 33

2.9 Questions 34

Chapter 3 Algebraic Operations (Multi–image point operations) 36

3.1 Image addition 36

3.2 Image subtraction (Differencing) 37

3.3 Image multiplication 38

3.4 Image division (Ratio) 39

3.5 Index derivation and supervised enhancement 42

3.5.1 Vegetation Indices 43

3.5.2 Iron oxide ratio index 43

3.5.3 TM clay (hydrated) mineral ratio index 44

3.6 Standardization and Logarithmic Residual 45

3.7 Simulated reflectance 46

3.7.1 Analysis of solar radiation balance and simulated irradiance 46

3.7.2 Simulated spectral reflectance image 47

3.7.3 Calculation of weights 49

3.7.4 Example: ATM simulated reflectance colour composite 50

3.7.5 Comparison with ratio and logarithmic residual techniques 51

3.8 Summary 52

3.9 Questions 53

Chapter 4 Filtering and neighbourhood processing 54

4.1 Fourier Transform: understanding filtering in image frequency 54

4.2 Concepts of convolution for image filtering 57

4.3 Low pass filters (smoothing) 58

4.4 High pass filters (edge enhancement) 63

4.4.1 Gradient filters 64

4.4.2 Laplacian filters 66

4.4.3 Edge sharpening filters 67

4.5 Local contrast enhancement 68

∗4.6 FFT selective and adaptive filtering 69

4.6.1 FFT selective filtering 71

4.6.2 FFT adaptive filtering 73

4.7 Summary 76

4.7 Questions 76

Chapter 5 RGB–IHS Transformation 78

5.1 Colour co–ordinate transformation 78

5.2 IHS decorrelation stretch 81

5.3 Direct decorrelation stretch technique 82

5.4 Hue RGB (HRGB) colour composites 86

∗5.5 Derivation of RGB–IHS and IHS–RGB transformation based on 3D geometry of the RGB colour cube 88

5.5.1 Derivation of RGB–IHS transformation 88

5.5.2 Derivation of IHS–RGB transformation 89

∗5.6 Mathematical proof of DDS and its properties 90

5.6.1 Mathematical proof of DDS 90

5.6.2 The properties of DDS 91

5.7 Summary 93

5.8 Questions 94

Chapter 6 Image Fusion Techniques 95

6.1 RGB–IHS transformation as a tool for data fusion 95

6.2 Brovey transform (intensity modulation) 96

6.3 Smoothing Filter based Intensity Modulation 98

6.3.1 The principle of SFIM 98

6.3.2 Merits and limitation of SFIM 99

6.3.3 An example of SFIM Pan–sharpen of Landsat–8 OLI image 101

6.4 Summary 103

6.5 Questions 103

Chapter 7 Principal Component Analysis (PCA) 104

7.1 Principle of the PCA 104

7.2 Principal component (PC) images and PC colour composition 107

7.3 Selective PCA (SPCA) for PC colour composition 110

7.3.1 Dimensionality and colour confusion reduction 111

7.3.2 Spectral contrast mapping 111

7.3.3 FPCS spectral contrast mapping 112

7.4 Decorrelation stretch 113

7.5 Physical property orientated coordinate transformation and Tasselled Cap Transformation 114

7.6 Statistic methods for band Selection 116

7.6.1 Review of Chavez′s and Sheffield′s methods 117

7.6.2 Index of three–dimensionality 117

7.7 Remarks 118

7.8 Questions 119

Chapter 8 Image classification 120

8.1 Approaches of statistical classification 120

8.2 Unsupervised classification (iterative clustering) 121

8.2.1 Iterative clustering algorithms 121

8.2.2 Feature space iterative clustering 123

8.2.3 Seed selection 124

8.2.4 Cluster splitting along PC1 125

8.3 Supervised classification 127

8.31 Generic algorithm of supervised classification 127

8.3.2 Spectral Angle Mapping Classification 127

8.4 Decision rules: dissimilarity functions 128

8.5 Post–classification processing: smoothing and accuracy assessment 130

8.5.1 Class smoothing process 130

8.5.2 Classification accuracy assessment 131

8.6 Summary 134

8.7 Questions 135

Chapter 9 Image Geometric Operations 136

9.1 Image geometric deformation 136

9.1.1 Platform flight coordinates, sensor status and imaging position 136

9.1.2 Earth rotation and curvature 138

9.2 Polynomial deformation model and image warping co–registration 139

9.2.1 Derivation of deformation model 140

9.2.2 Pixel DN re–sampling 141

9.3 Ground control point (GCP) selection and automation of image co–registration 142

9.3.1 Manual and semi–automatic GCP selection 142

9.3.2 Automatic image co–registration 143

9.4 Summary 144

9.5 Questions 145

Chapter 10 Introduction to interferometric synthetic aperture radar (InSAR) technique 146

10.1 The principle of a radar interferometer 146

10.2 Radar interferogram and DEM 148

10.3 Differential InSAR (DInSAR) and deformation measurement 151

10.4 Multi–temporal coherence image and random change detection 154

10.5 Spatial decorrelation and ratio coherence technique 157

10.6 Fringe smoothing filter 160

10.7 Summary 161

10.8 Questions 162

Chapter 11 Subpixel technology and its applications 163

11.1 Phase correlation algorithm 163

11.2 Phase correlation scanning for pixel–wise disparity estimation 168

11.2.1 Disparity estimation by phase correlation scanning 168

11.2.2 The median shift propagation (MSP) technique for disparity refinement 169

11.3 Pixel–wise image co–registration 171

11.3.1 Basic procedure of pixel–wise image co–registration using phase correlation 172

11.3.2 An example of pixel–wise image co–registration 172

11.3.3 Limitations 173

11.3.4 Pixel–wise image co–registration based SFIM Pan–sharpen 175

11.4 Very narrow baseline stereo matching and 3D data generation 177

11.4.1 The principle of stereo vision 177

11.4.2 Wide baseline vs. narrow baseline stereo 178

11.4.3 Narrow baseline stereo–matching using phase correlation 178

11.4.4 Accuracy assessment and application examples 179

11.5 Ground motion/deformation detection and estimation 183

11.5 Summary 185

Part II. Geographical Information Systems 187

Chapter 12 Geographical Information Systems 187

12.1 Introduction 187

12.2 Software tools 188

12.3 GIS, cartography and thematic mapping 188

12.4 Standards, interoperability and metadata 189

12.5 GIS and the Internet 190

Chapter 13 Data Models and Structures 191

13.1 Introducing spatial data in representing geographic features 191

13.2 How are spatial data different from other digital data? 191

13.3 Attributes and measurement scales 191

13.4 Fundamental data structures 192

13.5 Raster data 193

13.5.1 Data quantisation and storage 194

13.5.2 Spatial variability 195

13.5.3 Representing spatial relationships 196

13.5.4 The effect of resolution 196

13.5.5 Representing surface phenomena 197

13.6 Vector data 197

13.6.1 Vector Data Models 198

13.6.2 Representing logical relationships through geometry and feature definition 198

13.6.3 Extending the vector data model 203

13.6.4 Representing surfaces 206

13.7 Data conversion between models and structures 208

13.7.1 Vector to raster conversion (rasterisation) 209

13.7.2 Raster to vector conversion (vectorisation) 211

13.8 Summary 213

13.9 Questions 213

Chapter 14 Defining a coordinate space 214

14.1 Introduction 214

14.2 Datums and projections 214

14.2.1 Describing and measuring the earth 214

14.2.2 Measuring height: the geoid 216

14.2.3 Coordinate systems 216

14.2.4 Datums 217

14.2.5 Geometric distortions and projection models 218

14.2.6 Major Map Projections 221

14.2.7 Projection Specification 224

14.3 How coordinate information is stored and accessed 225

14.4 Selecting appropriate coordinate systems 226

14.5 Questions 227

Chapter 15 Operations 228

15.1 Introducing operations on spatial data 228

15.2 Map algebra concepts 229

15.2.1 Working with Null data 229

15.2.2 Logical and conditional processing 230

14.2.3 Other types of operator 230

15.3 Local operations 232

15.3.1 Primary operations 232

15.3.2 Unary operations 232

15.3.3 Binary operations 235

15.3.4 N–ary operations 237

15.4 Neighbourhood operations 237

15.4.1 Local neighbourhood 237

15.4.2 Extended neighbourhood 244

15.5 Vector equivalents to raster map algebra 245

15.6 Automating GIS functions 247

15.7 Summary 248

14.7 Questions 248

Chapter 16 Extracting information from point data: geostatistics 249

16.1 Introduction 249

16.2 Understanding the data 249

16.2.1 Histograms 249

15.2.2 Spatial autocorrelation 250

16.2.3 Variograms 251

16.2.4 Underlying Trends and Natural Barriers 253

16.3 Interpolation 254

16.3.1 Selecting sample size 254

16.3.2 Interpolation methods 255

16.3.3 Deterministic interpolators 256

16.3.4 Stochastic interpolators 261

16.4 Summary 264

16.5 Questions 264

Chapter 17 Representing and Exploiting Surfaces 266

17.1 Introduction 266

17.2 Sources and uses of surface data 266

17.2.1 Digital Elevation Models 266

17.2.2 Vector surfaces and objects 268

17.2.3 Uses of Surface Data 269

17.3 Visualising surfaces 270

17.3.1 Visualising in two dimensions 270

17.3.2 Visualising in three dimensions 273

17.4 Extracting surface parameters 277

17.4.1 Slope: gradient and aspect 277

17.4.2 Curvature 279

17.4.3 Surface topology: drainage networks and watersheds 282

17.4.4 Viewshed 285

17.4.5 Calculating volume 286

17.5 Summary 287

17.6 Questions 287

Chapter 18 Decision support and uncertainty 288

18.1 Introduction 288

18.2 Decision Support 288

18.3 Uncertainty 289

18.3.1 Criterion uncertainty 290

18.3.2 Threshold uncertainty 290

18.3.3 Decision rule uncertainty 291

18.4 Risk and hazard 291

18.5 Dealing with Uncertainty in GIS Based Spatial Analysis 292

18.5.1 Error Assessment (Criterion Uncertainty) 292

18.5.2 Fuzzy Membership (Threshold and Decision Rule Uncertainty) 293

18.5.3 Multi–Criteria Decision Making (Decision Rule Uncertainty) 294

18.5.4 Error Propagation and Sensitivity analysis (Decision Rule Uncertainty) 295

18.5.5 Result Validation (Decision Rule Uncertainty) 296

18.6 Summary 297

18.7 Key Questions 297

Chapter 19 Complex problems and multi–criterion evaluation 298

19.1 Introduction 298

19.2 Different approaches and models 299

19.2.1 Knowledge–driven (conceptual) 299

19.2.2 Data–driven (empirical) 299

19.2.3 Data–driven (neural–network) 300

19.3 Evaluation criteria 300

19.4 Deriving weighting coefficients 301

19.4.1 Rating 302

19.4.2 Ranking 302

19.4.3 Pairwise Comparison 303

19.5 Multi–criteria combination methods 305

19.5.1 Boolean logical combination 306

19.5.2 Index–overlay and algebraic combination 306

19.5.3 Weights of evidence modelling based on Bayesian Probability theory 306

19.5.4 Belief and Dempster–Shafer theory 308

19.5.5 Weighted factors in Linear Combination (WLC) 310

19.5.6 Fuzzy logic 313

19.5.7 Vectorial Fuzzy Modeling 314

19.6 Summary 316

19.7 Questions 316

Part III. Remote Sensing Applications 318

Chapter 20 Image Processing and GIS Operation Strategy 318

20.1 General image processing strategy 319

20.1.1 Preparation of basic working dataset 320

20.1.2 Image processing 323

20.1.3 Image interpretation and map composition 327

20.2 Remote sensing based GIS projects: from images to thematic mapping 329

20.3 An example of thematic mapping based on optimal visualisation and interpretation of multi–spectral satellite imagery 330

20.3.1 Background information 330

20.3.2 Image enhancement for visual observation 332

20.3.3 Data capture and image interpretation 333

20.3.4 Map composition 336

20.4 Summary 338

Chapter 21 Thematic Teaching Case Studies in SE Spain 339

21.1 Thematic information extraction (1): Gypsum natural outcrop mapping and quarry change assessment 339

21.1.1 Data preparation and general visualisation 339

21.1.2 Gypsum enhancement and extraction based on spectral analysis 340

21.1.3 Gypsum quarry changes during 1984–2000 343

21.1.4 Summary of the case study 345

21.1.5 Questions 345

21.2 Thematic information extraction (2): Spectral enhancement and mineral mapping of epithermal gold alteration, and iron–ore deposits in ferroan dolomite 346

21.2.1 Image datasets and data preparation 346

21.2.2 ASTER image processing and analysis for regional prospectivity 348

21.2.3 ATM image processing and analysis for target extraction 351

21.2.4 Summary 353

21.2.5 Questions 354

21.3 Remote sensing and GIS: evaluating vegetation and landuse change in the Nijar Basin, SE Spain. 357

21.3.1 Introduction 357

21.3.2 Data Preparation 359

21.3.3 Highlighting vegetation 359

21.3.4 Highlighting plastic greenhouses 361

21.3.5 Identifying change between different dates of observation 364

21.3.6 Summary 366

21.3.7 Questions 367

21.3.8 References 367

21.4 Applied remote sensing and GIS: a combined interpretive tool for regional tectonics, drainage and water resources in the Andarax basin 368

21.4.1 Introduction 368

21.4.2 Geological & hydrological setting 368

21.4.3 Case study objectives 369

21.4.5 Landuse and vegetation 371

21.4.6 Lithological enhancement and discrimination 374

21.4.7 Structural enhancement and interpretation 378

21.4.8 Summary 384

21.4.8 Questions 385

21.4.9 References 385

Chapter 22 Research Case Studies 386

22.1 Vegetation change in the Three Parallel Rivers region, Yunnan Province, China 386

22.1.1 Introduction 386

22.1.2 The study area and data 386

22.1.3 NDVI Difference Red, Green and Intensity (NDVI–D–RGI) composite 387

22.1.4 Data processing 389

22.1.5 Interpretation of regional vegetation changes 391

22.1.6 Summary 396

22.1.7 References 397

22.2 GIS modelling of earthquake damage zones using satellite imagery and DEM data 399

22.2.1 Introduction 399

22.2.2 The models 403

22.2.3 Derivation of input variables 405

22.2.4 Earthquake Damage Zone Modelling and Assessment 417

22.2.5 Summary 422

22.2.6 References 423

22.3 Predicting landslides using fuzzy geohazard mapping; an example from Piemonte, north–west Italy 426

22.3.1 Introduction 426

22.3.2 The study area 427

22.3.3 A holistic GIS based approach to landslide hazard assessment 431

22.3.4 Summary 436

22.3.6 Questions 437

22.3.7 References 437

22.4 Land surface change detection in a desert area in Algeria using multi–temporal ERS SAR coherence images 441

22.4.1 The study area 441

22.4.2 Coherence image processing and evaluation 442

22.4.3 Image visualisation & interpretation for change detection 443

22.4.4 Summary 448

22.4.5 References 449

Chapter 23 Industrial Case Studies 451

23.1 Multi–criteria assessment of mineral prospectivity, in SE Greenland 451

23.1.1 Introduction and objectives 451

23.1.2 Area description 451

23.1.3 Litho–tectonic context why the project s concept works 453

23.1.4 Mineral Deposit Types Evaluated 454

23.1.5 Data preparation 454

23.1.6 Multi–criteria spatial modeling 464

23.1.7 Summary 467

23.1.8 Questions 468

23.1.9 Acknowledgements 468

23.1.10 References 468

23.2 Water resource exploration in Somalia 470

23.2.1 Introduction 470

23.2.2 Data Preparation 471

23.2.3 Preliminary geological enhancements and target area identification 472

23.2.4 Discrimination potential aquifer lithologies using ASTER spectral indices 476

23.2.5 Summary 482

23.2.6 Questions 483

23.2.7 References 483

Part 4 Summary 484

Chapter 24 Concluding remarks 484

24.1 Image processing 484

24.2 Geographic Information Systems 487

24.3 Final remarks 491

Appendix A: Imaging Sensor Systems and Remote Sensing Satellites 492

A.1 Multi–spectral sensing 492

A.2 Broad band multi–spectral sensors 494

A.2.1 Digital camera 495

A.2.2 Across–track Mechanical Scanner 495

A.2.3 Along–track Push–broom Scanner 496

A.3 Thermal sensing and thermal infrared (TIR) sensors 497

A.4 Hyper–spectral sensors (Imaging spectrometers) 499

A.5 Passive microwave sensors 499

A.6 Active sensing: Synthetic Aperture Radar (SAR) imaging systems 500

Appendix B: Online resources for information, software and data 508

B.1 Software proprietary, low cost and free (shareware): 508

B.2 Information and technical information on standards, best practice, formats, techniques and various publications: 508

B.3 Data sources including online satellite imagery from major suppliers, DEM data plus GIS maps and data of all kinds: 509

References 510

General References 510

Image Processing 510

GIS 510

Remote Sensing 511

Part–I References and Further Reading 511

Part–II References and Further Reading 517

Index 524

Jian Guo Liu  received a Ph.D. in 1991 in remote sensing and image processing from Imperial College London, UK and an M.Sc. in 1982 in remote sensing and geology from China University of Geosciences, Beijing, China. He is a Reader in remote sensing in the Department of Earth Science and Engineering, Imperial College London. His current research activities include: sub–pixel technology for image registration, DEM generation and change detection; image processing techniques for data fusion, filtering and InSAR; and GIS multi–data modelling for geohazard studies.

Philippa J Mason completed a BSc in Geology at Southampton University in 1987, an MSc in Remote Sensing at University College London in 1993 and a PhD in 1998 at Imperial College London. She is a lecturer in remote sensing & GIS at Imperial College London and a consultant in geological remote sensing and image interpretation. Her research interests include the application of geospatial sciences to geohazards, tectonic geomorphology, spectral geology and mineral exploration.