Essential Image Processing and GIS for Remote Sensing. Techniques and Applications

472 Pages·2009·30.27 MB·English

by  Liu J.G.

Image processing and GIS for remote Sensing

Image processing
and GIS for remote
Sensing
Techniques and applications
Jian Guo Liu and philippa J. Mason
Department of Earth Science and Engineering,
Imperial College London
Second edItIon
This edition first published 2016 © 2016 by John Wiley & Sons, Ltd.
Registered Office
John Wiley & Sons, Ltd., The atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Offices
9600 Garsington Road, Oxford, OX4 2DQ, UK
The atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
111 River Street, Hoboken, NJ 07030‐5774, USa
For details of our global editorial offices, for customer services and for information about how
to apply for permission to reuse the copyright material in this book please see our website at
www.wiley.com/wiley‐blackwell.
The right of the author to be identified as the author of this work has been asserted in accordance with the
UK Copyright, Designs and Patents act 1988.
all rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
except as permitted by the UK Copyright, Designs and Patents act 1988, without the prior permission of
the publisher.
Designations used by companies to distinguish their products are often claimed as trademarks. all brand
names and product names used in this book are trade names, service marks, trademarks or registered
trademarks of their respective owners. The publisher is not associated with any product or vendor
mentioned in this book.
Limit of Liability/Disclaimer of Warranty: While the publisher and author(s) have used their best efforts in
preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is not
engaged in rendering professional services and neither the publisher nor the author shall be liable for
damages arising herefrom. If professional advice or other expert assistance is required, the services of a
competent professional should be sought.
Library of Congress Cataloging‐in‐Publication data applied for
ISBN: 9781118724200
a catalogue record for this book is available from the British Library.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not
be available in electronic books.
Cover image: The cover image is a 3D perspective view of a Landsat 8 OLI (Operational Land Imager)
image, draped over a Digital Elevation Model (SRTM3 1-arc-second 30 m), displaying rocks of Proterozoic
age outcropping over an area of the atlas Mountains of Morocco, which also host many mineral deposits.
The colour composite is composed of OLI bands 6-4-2 in RGB, with a DDS (Direct Decorrelation Stretch)
enhancement; bands which are in the Short-Wave InfraRed (6), Red (4) and Blue (2) spectral regions.
The bright green features in this image would therefore actually appear distinctly red to the naked eye.
Geologically, the beautiful and complex colour patterns are caused by folds and faults systems which
have deformed a series of different rock types exposed in the semi-arid environment of Morocco.
(Courtesy to NaSa/JPL and USGS).
Set in 8.5/12pt Meridien by SPi Global, Pondicherry, India
1 2016
Contents
Overview of the book, xi 3.6 Standardization and logarithmic residual, 29
3.7 Simulated reflectance, 29
3.7.1 Analysis of solar radiation balance and
Part I Image processing
simulated irradiance, 29
1 Digital image and display, 3 3.7.2 Simulated spectral reflectance image, 30
1.1 What is a digital image?, 3 3.7.3 Calculation of weights, 31
1.2 Digital image display, 4 3.7.4 Example: ATM simulated reflectance
1.2.1 Monochromatic display, 4 colour composite, 31
1.2.2 Tristimulus colour theory and RGB (red, 3.7.5 Comparison with ratio and logarithmic
green, blue) colour display, 5 residual techniques, 33
1.2.3 Pseudo‐colour display, 6 3.8 Summary, 33
1.3 Some key points, 8 3.9 Questions, 34
1.4 Questions, 8
4 Filtering and neighbourhood processing, 35
2 Point operations (contrast enhancement), 9 4.1 FT: Understanding filtering in image frequency, 35
2.1 Histogram modification and lookup table, 9 4.2 Concepts of convolution for image filtering, 37
2.2 Linear contrast enhancement (LCE), 11 4.3 Low pass filters (smoothing), 38
2.2.1 Derivation of a linear function 4.3.1 Gaussian filter, 39
from two points, 12 4.3.2 K nearest mean filter, 39
2.3 Logarithmic and exponential contrast 4.3.3 Median filter, 40
enhancement, 13 4.3.4 Adaptive median filter, 41
2.3.1 Logarithmic contrast enhancement, 13 4.3.5 K nearest median filter, 41
2.3.2 Exponential contrast enhancement, 14 4.3.6 Mode (majority) filter, 41
2.4 Histogram equalisation (HE), 14 4.3.7 Conditional smoothing filters, 41
2.5 Histogram matching (HM) and Gaussian stretch, 15 4.4 High pass filters (edge enhancement), 42
2.6 Balance contrast enhancement technique 4.4.1 Gradient filters, 43
(BCET), 16 4.4.2 Laplacian filters, 44
2.6.1 Derivation of coefficients a, b and c for a 4.4.3 Edge‐sharpening filters, 45
BCET parabolic function (Liu 1991), 17 4.5 Local contrast enhancement, 45
2.7 Clipping in contrast enhancement, 18 4.6 FFT selective and adaptive filtering, 46
2.8 Tips for interactive contrast enhancement, 18 4.6.1 FFT selective filtering, 47
2.9 Questions, 19 4.6.2 FFT adaptive filtering, 47
4.7 Summary, 52
3 Algebraic operations (multi‐image point
4.8 Questions, 52
operations), 21
3.1 Image addition, 21 5 RGB‐IHS transformation, 55
3.2 Image subtraction (differencing), 22 5.1 Colour co‐ordinate transformation, 55
3.3 Image multiplication, 22 5.2 IHS de‐correlation stretch, 57
3.4 Image division (ratio), 22 5.3 Direct de‐correlation stretch technique, 58
3.5 Index derivation and supervised enhancement, 26 5.4 Hue RGB colour composites, 60
3.5.1 Vegetation indices, 26 5.5 Derivation of RGB‐IHS and IHS‐RGB
3.5.2 Iron oxide ratio index, 27 transformation based on 3D geometry of the
3.5.3 T M clay (hydrated) mineral ratio index, 27 RGB colour cube, 63
v
vi Contents
5.5.1 Derivation of RGB‐IHS transformation, 63 8.2.3 Seed selection, 94
5.5.2 Derivation of IHS‐RGB transformation, 64 8.2.4 Cluster splitting along PC1, 95
5.6 Mathematical proof of DDS and its properties, 65 8.3 Supervised classification, 96
5.6.1 Mathematical proof of DDS, 65 8.3.1 Generic algorithm of supervised
5.6.2 The properties of DDS, 65 classification, 96
5.7 Summary, 67 8.3.2 Spectral angle mapping classification, 96
5.8 Questions, 67 8.4 Decision rules: Dissimilarity functions, 97
8.4.1 Box classifier, 97
6 Image fusion techniques, 69 8.4.2 Euclidean distance: Simplified
6.1 R GB‐IHS transformation as a tool for data maximum likelihood, 97
fusion, 69 8.4.3 Maximum likelihood, 97
6.2 B rovey transform (intensity modulation), 71 8.4.4 Optimal multiple point re‐assignment
6.3 S moothing filter‐based intensity modulation, 71 (OMPR), 98
6.3.1 The principle of SFIM, 71 8.5 Post‐classification processing: Smoothing and
6.3.2 Merits and limitations of SFIM, 73 accuracy assessment, 98
6.3.3 An example of SFIM pan‐sharpen 8.5.1 Class smoothing process, 98
of Landsat 8 OLI image, 75 8.5.2 Classification accuracy assessment, 98
6.4 Summary, 75 8.6 Summary, 101
6.5 Questions, 75 8.7 Questions, 101
9 Image geometric operations, 103
7 Principal component analysis, 77
9.1 Image geometric deformation, 103
7.1 Principle of the PCA, 77
9.1.1 Platform flight coordinates, sensor status
7.2 PC images and PC colour composition, 79
and imaging position, 103
7.3 Selective PCA for PC colour composition, 82
9.1.2 Earth rotation and curvature, 105
7.3.1 Dimensionality and colour confusion
9.2 Polynomial deformation model and image
reduction, 83
warping co‐registration, 106
7.3.2 Spectral contrast mapping, 83
9.2.1 Derivation of deformation model, 106
7.3.3 FPCS spectral contrast mapping, 83
9.2.2 Pixel DN re‐sampling, 108
7.4 De‐correlation stretch, 84
9.3 GCP selection and automation of image
7.5 Physical property orientated coordinate
co‐registration, 109
transformation and tasselled cap
9.3.1 Manual and semi‐automatic GCP
transformation, 85
selection, 109
7.6 Statistical methods for band selection, 87
9.3.2 Automatic image co‐registration, 109
7.6.1 Review of Chavez’s and Sheffield’s
9.4 Summary, 110
methods, 87
9.5 Questions, 110
7.6.2 Index of three‐dimensionality, 88
7.7 Remarks, 88 10 Introduction to interferometric synthetic aperture
7.8 Questions, 89 radar technique, 113
10.1 The principle of a radar interferometer, 113
8 Image classification, 91 10.2 Radar interferogram and DEM, 115
8.1 A pproaches of statistical classification, 91 10.3 Differential InSAR and deformation
8.1.1 Unsupervised classification, 91 measurement, 117
8.1.2 Supervised classification, 91 10.4 Multi‐temporal coherence image and random
8.1.3 Classification processing change detection, 119
and implementation, 91 10.5 Spatial de‐correlation and ratio coherence
8.1.4 Summary of classification approaches, 92 technique, 121
8.2 Unsupervised classification (iterative clustering), 92 10.6 Fringe smoothing filter, 123
8.2.1 Iterative clustering algorithms, 92 10.7 Summary, 124
8.2.2 Feature space iterative clustering, 93 10.8 Questions, 125
Contents vii
11 Sub‐pixel technology and its applications, 127 13.5.4 The effect of resolution, 160
11.1 Phase correlation algorithm, 127 13.5.5 Representing surface phenomena, 160
11.2 PC scanning for pixel‐wise disparity 13.6 Vector data, 161
estimation, 132 13.6.1 Vector data models, 162
11.2.1 Disparity estimation by PC scanning, 132 13.6.2 Representing logical relationships
11.2.2 The median shift propagation through geometry and feature
technique for disparity refinement, 133 definition, 162
11.3 Pixel‐wise image co‐registration, 134 13.6.3 Extending the vector data
11.3.1 Basic procedure of pixel‐wise image model, 167
co‐registration using PC, 135 13.6.4 Representing surfaces, 168
11.3.2 An example of pixel‐wise image 13.7 Data conversion between models and
co‐registration, 135 structures, 171
11.3.3 Limitations, 136 13.7.1 Vector to raster conversion
11.3.4 Pixel‐wise image co‐registration‐based (rasterisation), 171
SFIM pan‐sharpen, 136 13.7.2 Raster to vector conversion
11.4 Very narrow‐baseline stereo matching and 3D (vectorisation), 173
data generation, 139 13.8 Summary, 174
11.4.1 The principle of stereo vision, 139 13.9 Questions, 175
11.4.2 Wide‐baseline vs. narrow‐baseline
14 Defining a coordinate space, 177
stereo, 140
14.1 Introduction, 177
11.4.3 Narrow‐baseline stereo matching
14.2 Datums and projections, 177
using PC, 140
14.2.1 Describing and measuring the
11.4.4 Accuracy assessment and application
earth, 177
examples, 141
14.2.2 Measuring height: The geoid, 179
11.5 Ground motion/deformation detection
14.2.3 Coordinate systems, 179
and estimation, 143
14.2.4 Datums, 180
11.6 Summary, 146
14.2.5 Geometric distortions and projection
models, 181
14.2.6 Major map projections, 183
Part II Geographical information systems
14.2.7 Projection specification, 187
12 Geographical information systems, 151 14.3 How coordinate information
12.1 Introduction, 151 is stored and accessed, 188
12.2 Software tools, 152 14.4 Selecting appropriate coordinate systems, 189
12.3 GIS, cartography and thematic mapping, 152 14.5 Questions, 189
12.4 Standards, inter‐operability and 15 Operations, 191
metadata, 153 15.1 Introducing operations on spatial data, 191
12.5 GIS and the internet, 154 15.2 Map algebra concepts, 192
13 Data models and structures, 155 15.2.1 Working with Null data, 192
13.1 Introducing spatial data in representing 15.2.2 Logical and conditional processing, 193
geographic features, 155 15.2.3 Other types of operator, 193
13.2 How are spatial data different from other 15.3 Local operations, 194
digital data?, 155 15.3.1 Primary operations, 195
13.3 Attributes and measurement scales, 156 15.3.2 Unary operations, 195
13.4 Fundamental data structures, 156 15.3.3 Binary operations, 198
13.5 Raster data, 157 15.3.4 N‐ary operations, 198
13.5.1 Data quantisation and storage, 157 15.4 Neighbourhood operations, 199
13.5.2 Spatial variability, 159 15.4.1 Local neighbourhood, 199
13.5.3 Representing spatial relationships, 159 15.4.2 Extended neighbourhood, 206
viii Contents
15.5 Vector equivalents to raster map algebra, 206 18.3 Uncertainty, 248
15.5.1 Buffers, 207 18.3.1 Criterion uncertainty, 248
15.5.2 Dissolve, 207 18.3.2 Threshold uncertainty, 249
15.5.3 Clipping, 208 18.3.3 Decision rule uncertainty, 249
15.5.4 Intersection, 208 18.4 Risk and hazard, 250
15.6 Automating GIS functions, 209 18.5 Dealing with uncertainty in GIS‐based spatial
15.7 Summary, 209 analysis, 250
15.8 Questions, 210 18.5.1 Error assessment (criterion
uncertainty), 251
16 Extracting information from point
18.5.2 Fuzzy membership (threshold and
data: Geostatistics, 211
decision rule uncertainty), 251
16.1 Introduction, 211
18.5.3 Multi‐criteria decision making
16.2 Understanding the data, 211
(decision rule uncertainty), 252
16.2.1 Histograms, 212
18.5.4 Error propagation and
16.2.2 Spatial auto‐correlation, 212
sensitivity analysis (decision rule
16.2.3 Variograms, 213
uncertainty), 253
16.2.4 Underlying trends and natural
18.5.5 Result validation (decision rule
barriers, 214
uncertainty), 254
16.3 Interpolation, 214
18.6 Summary, 254
16.3.1 Selecting sample size, 216
18.7 Questions, 255
16.3.2 Interpolation methods, 217
16.3.3 Deterministic interpolators, 217 19 Complex problems and multi‐criterion
16.3.4 Stochastic interpolators, 221 evaluation, 257
16.4 Summary, 224 19.1 Introduction, 257
16.5 Questions, 225 19.2 Different approaches and models, 258
19.2.1 Knowledge‐driven (conceptual), 258
17 Representing and exploiting surfaces, 227
19.2.2 Data‐driven (empirical), 258
17.1 Introduction, 227
19.2.3 Data‐driven (neural network), 259
17.2 Sources and uses of surface data, 227
19.3 Evaluation criteria, 259
17.2.1 Digital elevation models, 227
19.4 Deriving weighting coefficients, 260
17.2.2 Vector surfaces and objects, 229
19.4.1 Rating, 260
17.2.3 Uses of surface data, 230
19.4.2 Ranking, 261
17.3 Visualising surfaces, 230
19.4.3 Pairwise comparison, 261
17.3.1 Visualising in two dimensions, 231
19.5 Multi‐criterion combination methods, 263
17.3.2 Visualising in three dimensions, 233
19.5.1 Boolean logical combination, 263
17.4 Extracting surface parameters, 236
19.5.2 Index‐overlay and algebraic
17.4.1 Slope: Gradient and aspect, 236
combination, 264
17.4.2 Curvature, 238
19.5.3 Weights of evidence modelling based
17.4.3 Surface topology: Drainage networks
on Bayesian probability theory, 264
and watersheds, 241
19.5.4 Belief and Dempster‐Shafer
17.4.4 Viewshed, 242
Theory, 266
17.4.5 Calculating volume, 245
19.5.5 Weighted factors in linear combination
17.5 Summary, 245
(WLC), 267
17.6 Questions, 246
19.5.6 Fuzzy logic, 269
18 Decision support and uncertainty, 247 19.5.7 Vectorial fuzzy modelling, 270
18.1 Introduction, 247 19.6 Summary, 272
18.2 Decision support, 247 19.7 Questions, 272