Computer Processing of Remotely-Sensed Images E-Book

Paul M. Mather

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 1471

Veröffentlichungsjahr: 2011

Beschreibung

This fourth and full colour edition updates and expands a widely-used textbook aimed at advanced undergraduate and postgraduate students taking courses in remote sensing and GIS in Geography, Geology and Earth/Environmental Science departments. Existing material has been brought up to date and new material has been added. In particular, a new chapter, exploring the two-way links between remote sensing and environmental GIS, has been added. New and updated material includes: * A website at href="http://www.wiley.com/go/mather4">www.wiley.com/go/mather4 that provides access to an updated and expanded version of the MIPS image processing software for Microsoft Windows, PowerPoint slideshows of the figures from each chapter, and case studies, including full data sets, * Includes new chapter on Remote Sensing and Environmental GIS that provides insights into the ways in which remotely-sensed data can be used synergistically with other spatial data sets, including hydrogeological and archaeological applications, * New section on image processing from a computer science perspective presented in a non-technical way, including some remarks on statistics, * New material on image transforms, including the analysis of temporal change and data fusion techniques, * New material on image classification including decision trees, support vector machines and independent components analysis, and * Now in full colour throughout. This book provides the material required for a single semester course in Environmental Remote Sensing plus additional, more advanced, reading for students specialising in some aspect of the subject. It is written largely in non-technical language yet it provides insights into more advanced topics that some may consider too difficult for a non-mathematician to understand. The case studies available from the website are fully-documented research projects complete with original data sets. For readers who do not have access to commercial image processing software, MIPS provides a licence-free, intuitive and comprehensive alternative.

Leseprobe

Contents

Preface to the First Edition

Preface to the Second Edition

Preface to the Third Edition

Preface to the Fourth Edition

List of Examples

web page

1 Remote Sensing: Basic Principles

1.1 Introduction

1.2 Electromagnetic Radiation and Its Properties

1.3 Interaction with Earth-Surface Materials

1.4 Summary

2 Remote Sensing Platforms and Sensors

2.1 Introduction

2.2 Charahcteristics of Imaging Remote Sensing Instruments

2.3 Optical, Near-infrared and Thermal Imaging Sensors

2.4 Microwave Imaging Sensors

2.5 Summary

3 Hardware and Software Aspects of Digital Image Processing

3.1 Introduction

3.2 Properties of Digital Remote Sensing Data

3.3 Numerical Analysis and Software Accuracy

3.4 Some Remarks on Statistics

3.5 Summary

4 Preprocessing of Remotely-Sensed Data

4.1 Introduction

4.2 Cosmetic Operations

4.3 Geometric Correction and Registration

4.4 Atmospheric Correction

4.5 Illumination and View Angle Effects

4.6 Sensor Calibration

4.7 Terrain Effects

4.8 Summary

5 Image Enhancement Techniques

5.1 Introduction

5.2 Human Visual System

5.3 Contrast Enhancement

5.4 Pseudocolour Enhancement

5.5 Summary

6 Image Transforms

6.1 Introduction

6.2 Arithmetic Operations

6.3 Empirically Based Image Transforms

6.4 Principal Components Analysis

6.5 Hue-Saturation-Intensity (HSI) Transform

6.6 The Discrete Fourier Transform

6.7 The Discrete Wavelet Transform

6.8 Change Detection

6.9 Image Fusion

6.10 Summary

7 Filtering Techniques

7.1 Introduction

7.2 Spatial Domain Low-Pass (Smoothing) Filters

7.3 Spatial Domain High-Pass (Sharpening) Filters

7.4 Spatial Domain Edge Detectors

7.5 Frequency Domain Filters

7.6 Summary

8 Classification

8.1 Introduction

8.2 Geometrical Basis of Classification

8.3 Unsupervised Classification

8.4 Supervised Classification

8.5 Subpixel Classification Techniques

8.6 More Advanced Approaches to Image Classification

8.7 Incorporation of Non-spectral Features

8.8 Contextual Information

8.9 Feature Selection

8.10 Classification Accuracy

8.11 Summary

9 Advanced Topics

9.1 Introduction

9.2 SAR Interferometry

9.3 Imaging Spectroscopy

9.4 Lidar

9.5 Summary

10 Environmental Geographical Information Systems: A Remote Sensing Perspective

10.1 Introduction

10.2 Data Models, Data Structures and File Formats

10.3 Geodata Processing

10.4 Locational Analysis

10.5 Spatial Analysis

10.6 Environmental Modelling

10.7 Visualization

10.8 Multicriteria Decision Analysis of Groundwater Recharge Zones

10.9 Assessing Flash Flood Hazards by Classifying Wadi Deposits in Arid Environments

10.10 Remote Sensing and GIS in Archaeological Studies

Appendix A Accessing MIPS

Appendix B Getting Started with MIPS

Appendix C Description of Sample Image Datasets

Appendix D Acronyms and Abbreviations

References

Index

This edition first published 2011 © 2011 by John Wiley & Sons, Ltd.

Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientific, Technical and Medical business with Blackwell Publishing.

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. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. 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

Mather, Paul M.

Computer processing of remotely-sensed images: an introduction/Paul M. Mather and Magaly Koch. – 4th ed.

p. cm.

Includes index.

ISBN 978-0-470-74239-6 (cloth) – ISBN 978-0-470-74238-9 (pbk.) 1. Remote sensing – Data processing.

2. Remote-sensing images. 3. Image processing – Digital techniques. I. Koch, Magaly. II. Title.

G70.4.M38 2011

621.36’78028566 – dc22

2010018060

A catalogue record for this book is available from the British Library.

This book is published in the following electronic formats: ePDF 9780470666500; Wiley Online Library 9780470666517

Set in 10/12pt Times by Laserwords Private Limited, Chennai, India

First Impression 2011

Preface to the First Edition

Environmental remote sensing is the measurement, from a distance, of the spectral features of the Earth’s surface and atmosphere. These measurements are normally made by instruments carried by satellites or aircraft, and are used to infer the nature and characteristics of the land or sea surface, or of the atmosphere, at the time of observation. The successful application of remote sensing techniques to particular problems, whether they be geographical, geological, oceanographic or cartographic, requires knowledge and skills drawn from several areas of science. An understanding of the way in which remotely sensed data are acquired by a sensor mounted onboard an aircraft or satellite needs a basic knowledge of the physics involved, in particular environmental physics and optics. The use of remotely-sensed data, which are inherently digital, demands a degree of mathematical and statistical skill plus some familiarity with digital computers and their operation. A high level of competence in the field in which the remotely-sensed data are to be used is essential if full use of the information contained in those data is to be made. The term ’remote-sensing specialist' is thus, apparently, a contradiction in terms for a remote-sensing scientist must possess a broad range of expertise across a variety of disciplines. While it is, of course, possible to specialise in some particular aspect of remote sensing, it is difficult to cut oneself off from the essential multidisciplinary nature of the subject.

This book is concerned with one specialised area of remote sensing, that of digital image processing of remotely sensed data but, as we have seen, this topic cannot be treated in isolation and so Chapter 1 covers in an introductory fashion the physical principles of remote sensing. Satellite platforms currently or recently in use, as well as those proposed for the near future, are described in Chapter 2, which also contains a description of the nature and sources of remotely sensed data. The characteristics of digital computers as they relate to the processing of remotely-sensed image data is the subject of Chapter 3. The remaining five chapters cover particular topics within the general field of the processing of remotely-sensed data in the form of digital images, and their application to a range of problems drawn from the Earth and environmental sciences. Chapters 1–3 can be considered to form the introduction to the material treated in later chapters.

The audience for this book is perceived as consisting of undergraduates taking advanced options in remote sensing in universities and colleges as part of a first degree course in geography, geology, botany, environmental science, civil engineering or agricultural science, together with postgraduate students following taught Masters courses in remote sensing. In addition, postgraduate research students and other research workers whose studies involve the use of remotely-sensed images can use this book as an introduction to the digital processing of such data. Readers whose main scientific interests lie elsewhere might find here a general survey of this relatively new and rapidly-developing area of science and technology. The nature of the intended audience requires that the formal presentation is kept to a level that is intelligible to those who do not have the benefit of a degree in mathematics, physics, computer science or engineering. This is not a research monograph, complete in every detail and pushing out to the frontiers of knowledge. Rather, it is a relatively gentle introduction to a subject which can, at first sight, appear to be overwhelming to those lacking mathematical sophistication, statistical cunning or computational genius. As such it relies to some extent on verbal rather than numerical expression of ideas and concepts. The author’s intention is to provide the foundations upon which readers may build their knowledge of the more complex and detailed aspects of the use of remote-sensing techniques in their own subject rather than add to the already extensive literature which caters for a mathematically-orientated readership. Because of the multidisciplinary nature of the intended audience, and since the book is primarily concerned with techniques, the examples have been kept simple, and do not assume any specialist knowledge of geology, ecology, oceanography or other branch of Earth science. It is expected that the reader is capable of working out potential applications in his or her own field, or of following up the references given here.

It is assumed that most readers will have access to a digital image processing system, either within their own department or institution, or at a regional or national remote sensing centre. Such processors normally have a built-in software package containing programs to carry out most, if not all, of the operations described in this book. It is hoped that the material presented here will provide such readers with the background necessary to make sensible use of the facilities at their disposal. Enough detail is provided, however, to allow interested readers to develop computer programs to implement their own ideas or to modify the operation of a standard package. Such ventures are to be encouraged, for software skills are an important part of the remote sensing scientist’s training. Furthermore, the development and testing of individual ideas and conjectures provides the opportunity for experiment and innovation, which is to be preferred to the routine use of available software. It is my contention that solutions to problems are not always to be found in the user manual of a standard software package.

I owe a great deal to many people who have helped or encouraged me in the writing of this book. Michael Coombes of John Wiley & Sons, Ltd. took the risk of asking me to embark upon the venture, and has proved a reliable and sympathetic source of guidance as well as a model of patience. The Geography Department, University of Nottingham, kindly allowed me to use the facilities of the Remote Sensing Unit. I am grateful also to Jim Cahill for many helpful comments on early drafts, to Michael Steven for reading part of Chapter 1 and for providing advice on some diagrams, and to Sally Ashford for giving a student’s view and to George Korybut-Daszkiewicz for his assistance with some of the photography. An anonymous referee made many useful suggestions. My children deserve a mention (my evenings on the word-processor robbed them of the chance to play their favourite computer games) as does my wife for tolerating me. The contribution of the University of Nottingham and the Shell Exploration and Development Company to the replacement of an ageing PDP11 computer by a VAX 11/730-based image processing system allowed the continuation of remote sensing activities in the university and, consequently, the successful completion of this book. Many of the ideas presented here are the result of the development of the image processing software system now in use at the University of Nottingham and the teaching of advanced undergraduate and Masters degree courses. I am also grateful to Mr J. Winn, Chief Technician, Mr. C. Lewis and Miss E. Watts, Cartographic Unit and Mr M.A. Evans, Photographic Unit, Geography Department, University of Nottingham, for their invaluable and always friendly assistance in the production of the photographs and diagrams. None of those mentioned can be held responsible for any errors or misrepresentations that might be present in this book; it is the author’s prerogative to accept liability for these.

Paul M. Mather, March 1987

Remote Sensing Unit, Department of Geography, The University of Nottingham, Nottingham NG7 2RD, UK

Preface to the Second Edition

Many things have changed since the first edition of this book was written, more than 10 years ago. The increasing emphasis on scientific rigour in remote sensing (or Earth observation by remote sensing, as it is now known), the rise of interest in global monitoring and large-scale climate modelling, the increasing number of satellite-borne sensors in orbit, the development of Geographical Information Systems (GISs) technology, and the expansion in the number of taught Masters courses in GIS and remote sensing are all noteworthy developments. Perhaps the most significant single change in the world of remote sensing over the past decade has been the rapid increase in and the significantly reduced cost of computing power and software available to students and researchers alike, which allows them to deal with growing volumes of data and more sophisticated and demanding processing tools. In 1987 the level of computing power available to researchers was minute in comparison with that which is readily available today. I wrote the first edition of this book using a BBC Model B computer, which had 32 kb of memory, 100 kb diskettes and a processor which would barely run a modern refrigerator. Now I am using a 266 MHz Pentium II with 64 Mb of memory and a 2.1 Gb disk. It has a word processor that corrects my spelling mistakes (though its grammar checking can be infuriating). I can connect from my home to the University of Nottingham computers by fibre optic cable and run advanced software packages. The cost of this computer is about 1% of that of the VAX 11/730 that is mentioned in the preface to the first edition of this book.

Although the basic structure of the book remains largely unaltered, I have taken the opportunity to revise all of the chapters to bring them up to date, as well as to add some new material, to delete obsolescent and uninteresting paragraphs, and to revise some infelicitous and unintelligible passages. For example, Chapter 4 now contains new sections covering sensor calibration, plus radiometric and topographic correction. The use of artificial neural networks in image classification has grown considerably in the years since 1987, and a new section on this topic is added to Chapter 8, which also covers other recent developments in pattern recognition and methods of estimating Earth surface properties. Chapter 3, which provides a survey of computer hardware and software, has been almost completely re-written. In Chapter 2 I have tried to give a brief overview of a range of present and past sensor systems but have not attempted to give a full summary of every sensor, because details of new developments are now readily available via the Internet. I doubt whether anyone would read this book simply because of its coverage of details of individual sensors.

Other chapters are less significantly affected by recent research as they are concerned with the basics of image processing (filtering, enhancement and image transforms), details of which have not changed much since 1987, though I have added new references and attempted to improve the presentation. I have, however, resisted the temptation to write a new chapter on GIS, largely because there are several good books on this topic that are widely accessible (for example Bonham-Carter (1994) and McGuire et al. (1991)), but also because I feel that this book is primarily about image processing. The addition of a chapter on GIS would neither do justice to that subject nor enhance the reader’s understanding of digital processing techniques. However, I have made reference to GIS and spatial databases at a number of appropriate points in the text. My omission of a survey of GIS techniques does not imply that I consider digital image processing to be a ‘stand-alone' topic. Clearly, there are significant benefits to be derived from the use of spatial data of all kinds within an integrated environment, and this point is emphasized in a number of places in this book. I have added a significant number of new references to each of the chapters, in the hope that readers might be encouraged to enjoy the comforts of his or her local library.

I have added a number of ‘self-assessment' questions at the end of each chapter. These questions are not intended to constitute a sample examination paper, nor do they provide a checklist of ’important' topics (the implication being that the other topics covered in the book are unimportant). They are simply a random set of questions–if you can answer them then you probably understand the contents of the chapter. Readers should use the Mather’s Image Processing System (MIPS) software described in Appendices A and B to try out the methods mentioned in these questions. Data sets are also available on the accompanying CD, and are described in Appendix C.

Perhaps the most significant innovation that this book offers is the provision of a CD containing software and images. I am not a mathematician, and so I learn by trying out ideas rather than exclusively by reading or listening. I learn new methods by writing computer programs and applying them to various data sets. I am including a small selection of the many programs that I have produced over the past 30 years, in the hope that others may find them useful. These programs are described in Appendix B. I have been teaching a course on remote sensing for the last 14 years. When this course began there were no software packages available, so I wrote my own (my students will remember NIPS, the Nottingham Image Processing System, with varying degrees of hostility). I have completely re-written and extended NIPS so that it now runs under Microsoft Windows 95. I have renamed it to MIPS, which is rather an unimaginative name, but is nevertheless pithy. It is described in Appendix A. Many of the procedures described in this book are implemented in MIPS, and I encourage readers to try out the methods discussed in each chapter. It is only by experimenting with these methods, using a range of images, that you will learn how they work in practice. MIPS was developed on an old 486-based machine with 12 Mb of RAM and a 200 Mb disk, so it should run on most PCs available in today’s impoverished universities and colleges. MIPS is not a commercial system, and should be used only for familiarisation before the reader moves on to the software behemoths that are so readily available for both PCs and UNIX workstations. Comments and suggestions for improving MIPS are welcome (preferably by email) though I warn readers that I cannot offer an advisory service nor assist in research planning!

Appendix C contains a number of Landsat, SPOT, AVHRR and RADARSAT images, mainly extracts of size 512 512 pixels. I am grateful to the copyright owners for permission to use these data sets. The images can be used by the reader to gain practical knowledge and experience of image processing operations. Many university libraries contain map collections, and I have given sufficient details of each image to allow the reader to locate appropriate maps and other back-up material that will help in the interpretation of the features shown on the images.

The audience for this book is seen to be advanced undergraduate and Masters students, as was the case in 1987. It is very easy to forget that today’s student of remote sensing and image processing is starting from the same level of background knowledge as his or her predecessors in the 1980s. Consequently, I have tried to restrain myself from including details of every technique that is mentioned in the literature. This is not a research monograph or a literature survey, nor is it primarily an exercise in self-indulgence and so some restriction on the level and scope of the coverage provided is essential if the reader is not to be overwhelmed with detail and thus discouraged from investigating further. Nevertheless, I have tried to provide references on more advanced subjects for the interested reader to follow up. The volume of published material in the field of remote sensing is now very considerable, and a full survey of the literature of the last 20 years or so would be both unrewarding and tedious. In any case, online searches of library catalogues and databases are now available from networked computers. Readers should, however, note that this book provides them only with a background introduction–successful project work will require a few visits to the library to peruse recent publications, as well as practical experience of image processing.

I am most grateful for comments from readers, a number of whom have written to me, mainly to offer useful suggestions. The new edition has, I hope, benefited from these ideas. Over the past years, I have been fortunate enough to act as supervisor to a number of postgraduate research students from various countries around the world. Their enthusiasm and commitment to research have always been a factor in maintaining my own level of interest, and I take this opportunity to express my gratitude to all of them. My friends and colleagues in the Remote Sensing Society, especially Jim Young, Robin Vaughan, Arthur Cracknell, Don Hardy and Karen Korzeniewski, have always been helpful and supportive. Discussions with many people, including Mike Barnsley, Giles Foody and Robert Gurney, have added to my knowledge and awareness of key issues in remote sensing. I also acknowledge with gratitude the help given by Dr Magaly Koch, Remote Sensing Center, Boston University, who has tested several of the procedures reported in this book and included on the CD. Her careful and thoughtful advice, support and encouragement have kept me from straying too far from reality on many occasions. My colleagues in the School of Geography in the University of Nottingham continue to provide a friendly and productive working environment, and have been known occasionally to laugh at some of my jokes. Thanks especially to Chris Lewis and Elaine Watts for helping to sort out the diagrams for the new edition, and to Dee Omar for his patient assistance and support. Michael McCullagh has been very helpful, and has provided a lot of invaluable assistance. The staff of John Wiley & Sons has been extremely supportive, as always. Finally, my wife Rosalind deserves considerable credit for the production of this book, as she has quietly undertaken many of the tasks that, in fairness, I should have carried out during the many evenings and weekends that I have spent in front of the computer. Moreover, she has never complained about the chaotic state of our dining room, nor about the intrusive sound of Wagner’s music dramas. There are many people, in many places, who have helped or assisted me; it is impossible to name all of them, but I am nevertheless grateful. Naturally, I take full responsibility for all errors and omissions.

Paul M. Mather

Nottingham, June 1998

Preface to the Third Edition

In the summer of 2001 I was asked by Lyn Roberts of John Wiley & Sons, Ltd. to prepare a new edition of this book. Only minor updates would be needed, I was told, so I agreed. A few weeks later was presented with the results of a survey of the opinions of the ‘great and the good’ as to what should be included in and what should be excluded from the new edition. You are holding the result in your hands. The ‘minor updates’ turned into two new chapters (a short one on computer basics, replacing the old Chapter 3, and a lengthier one on the advanced topics of interferometry, imaging spectroscopy and lidar, making a new Chapter 9) plus substantial revisions of the other chapters. In addition, I felt that development of the MIPS software would be valuable to readers who did not have access to commercial remote sensing systems. Again, I responded to requests from postgraduate students to include various modules that they considered essential, and the result is a Windows-based package of 90000+ lines of code.

Despite these updates and extensions both to the text of the book and the accompanying software, my target audience is still the advanced undergraduate taking a course in environmental remote sensing. I have tried to introduce each topic at a level that is accessible to the reader who is just becoming aware of the delights of image processing while, at the same time, making the reasonable assumption that my readers are, typically, enthusiastic, aware and intelligent, and wish to go beyond the basics. In order to accommodate this desire to read widely, I have included an extensive reference list. I am aware, too, that this book is used widely by students taking Masters level courses. Some of the more advanced material, for example in Chapters 6, 8 and 9, is meant for them; for example the new material on wavelets and developments in principal components analysis may stimulate Masters students to explore these new methods in their dissertation work. The first three chapters should provide a basic introduction to the background of remote sensing and image processing; Chapters 4–8 introduce essential ideas (noting the remark above concerning parts of Chapters 6 and 8), while Chapter 9 is really for the postgraduate or the specialist undergraduate.

I am a firm believer in learning by doing. Reading is not a complete substitute for practical experience of the use of image processing techniques applied to real data that relates to real problems. For most people, interest lies in the meaning of the results of an operation in terms of the information that is conveyed about a problem rather than in probing the more arcane details of particular methods, though for others it is the techniques themselves that fascinate. The level of mathematical explanation has therefore been kept to a minimum and I have attempted to use an approach involving examples, metaphors and verbal explanation. In particular, I have introduced a number of examples, separate from the main text, which should help the reader to interpret image-processing techniques in terms of real-world applications.

Many of these examples make use of the MIPS software that is provided on the CD that accompanies this book. MIPS has grown somewhat since 1999, when the second edition of this book was published. It has a new user interface, and is able to handle images of any size in 8-, 16- or 32-bit representation. A number of new features have been added, and it is now capable of providing access to many of the techniques discussed in this book. I would appreciate reports from readers of any difficulties they experience with MIPS, and I will maintain a web site from which updates and corrections can be downloaded. The URL of this web site, and my email address, can be found in the file contactme.txt which is located in the root directory of the CD.

Many of the ideas in this book have come from my postgraduate students. Over the past few years, I have supervised a number of outstanding research students, whose work has kept me up to date with new developments. In particular, I would like to thank Carlos Vieira, Brandt Tso, Taskin Kavzoglu, Premelatha Balan, Mahesh Pal, Juazir Hamid, Halmi Kamarrudin and Helmi Shafri for their tolerance and good nature. Students attending my Masters classes in digital image processing have also provided frank and valuable feedback. I would also like to acknowledge the valuable assistance provided by Rosemary Hoole and Karen Laughton of the School of Geography, University of Nottingham. The help of Dr Koch of Boston University, who made many useful comments on the manuscript and the MIPS software as they have progressed, is also gratefully acknowledged, as is the kindness of Professor J. Gumuzzio and his group at the Autonomous University of Madrid in allowing me access to DAIS images of their La Mancha study site. The DAIS data were recorded by DLR in the frame of the EC funded programme ‘Access to Research Infrastructures’, project ‘DAIS/ROSIS–Imaging Spectrometers at DLR’, Co. Ro. HPRI-CT-1999-0075. Dr Koch has also provided a set of four advanced examples, which can be found in the Examples folder of the accompanying CD. I am very grateful to her for this contribution, which I am sure significantly enhances the value of the book. My wife continues to tolerate what she quietly considers to be my over-ambitious literary activities, as well as my predilection for the very loudest bits of Mahler, Berlioz, Wagner and others. Colleagues and students of the School of Geography, University of Nottingham, have helped in many ways, not least by humouring me. Finally, I would like to thank Lyn Roberts, Kiely Larkins, and the staff of John Wiley who have helped to make this third edition a reality, and showed infinite patience and tolerance.

A book without errors is either trivial or guided by a divine hand. I can’t believe that this book is in the latter category, and it is possible that it isn’t in the former. I hope that the errors that you do find, for which I take full responsibility, are not too serious and that you will report them to me.

Paul M. Mather

Nottingham, August 2003

Preface to the Fourth Edition

Almost 25 years have passed since the publication of the first edition of this book. Back in 1987, image processing and geographical information systems (GIS) required dedicated computers that lived in air-conditioned rooms. At that time, I had access to a VAX 11/730, which had about the same computing power as a mobile phone of today. We had four megabytes of RAM and it took 12 hours to carry out a maximum likelihood classification on a 512 × 512 Landsat four-band MSS image. Imagery with the highest multispectral spatial resolution (20 m) was acquired by the SPOT-1 satellite, and there was not a great deal of software to process it. In those days an ability to write your own programmes was essential.

Over those 25 years, new companies, new products and new applications have come into being. Integrated software packages running on fast desktop machines, processing data that may have up to 256 bands, or a spatial resolution of 50 cm, with stereo capability for DEM production, are all seen as a natural if rapid development over the period 1987–2010. Yet the basic principles of image processing remain the same, and so does the structure of this book. However, a new chapter, number ten, has been added. It covers the topic of Environmental Geographical Information Systems and Remote Sensing. I wrote in the preface to the second edition (1999) that I had resisted the temptation to include a chapter on GIS because there were several good GIS texts already available. However, developments in the fields of image processing and GIS mean that nowadays no environmental remote sensing course is complete without some reference to environmental GIS, and no environmental GIS course can be considered adequate without a discussion of the uses of remotely sensed data. In fact, over the past two or three years there has been a ‘coming together’ of GIS, remote sensing and photogrammetry. Nowadays most software vendors describe their products not as ‘Image Processing for Remote Sensing’ but as geospatial, integrated, and so on. They can interact with GIS either directly or through a common data format. This book has, therefore, to include material on GIS from a remote sensing point of view. Being aware of my inadequacies in the field of GIS, I asked Dr Magaly Koch of Boston University’s Centre for Remote Sensing to help with the new Chapter 10 and with other chapters, as she has the benefit of considerable practical experience in the areas of remote sensing and GIS. I am very grateful to her for accepting this invitation and for making a valuable contribution to the book as a whole.

Perhaps the most important change in the appearance of the book is the introduction of colour illustrations and tables. This development is of tremendous value in a book on image processing. The use of colour supplements is a half-way house, and I always found it more trouble than it was worth. I am grateful to the publishers, and in particular to Fiona Woods, for negotiating a new full colour edition. In retrospect, trying to describe the effects of, for example, a decorrelation stretch while using only greyscale illustrations would have been a challenge to any of the great literary figures of the past, and I am sure that my powers of descriptive writing are but a shadow of theirs.

As far as other changes are concerned, I have rewritten most of Chapter 3. In the third edition Chapter 3 described the workings of the MIPS software package, which was included on a CD that accompanied the book. MIPS is still available, and can now be downloaded from the publisher’s web site (www.wiley.com/go/mather4). Chapter 3 now deals with aspects of computing and statistics which should be appreciated and understood by all users of computers. In particular, I feel strongly that users of remote sensing and GIS software packages should have a good understanding of the way their favoured package handles data; many people have been surprised to find that sometimes software does not always do what it says on the box. In comparison, economists and computer scientists have tested spreadsheet software such as Excel almost to destruction (McCullough, 1998, 1999). It is particularly important for the user to be aware of the nature and type of data being processed, as 8-, 10-, 12- and 16-bit data values are now in widespread use. Yet most graphics cards still limit displayed data to 8 bits per primary colour (red, green and blue) and the way that data are converted from their original format (8-, 10-, 11- or 16-bit) to fit the requirements of the graphics hardware is a matter of considerable interest, as the appearance of the image depends to a greater or lesser extent on this conversion process.

Other topics that are new to the fourth edition (apart from Chapters 3 and 10) include: updated information on new satellites and sensors (Chapter 2), new sections on change detection, and image fusion (Chapter 6), a description of the Frost filter (Chapter 7), a new section on independent components analysis (Chapter 8) plus descriptions of recent developments in image classification (support vector machines and decision trees, object-oriented classification and bagging and boosting). Other chapters have been updated, with old references removed and replaced by new ones, though several key papers from the 1970s to 1980s remain in the bibliography where they provide an uniquely intelligible description of a specific topic.

Some may query the size of the bibliography. However, modern bachelors’ and Masters’ programmes are more focused on assessed project work than was the case in the 1980s, and the key papers in the recent literature are catalogued in the following pages in order to provide guidance to students who are undertaking projects. I have tried to ensure that most journal paper references postdate 1994, because many electronically accessible journals are available only from the mid-1990s on, and I know that both students and researchers like to read their favourite journals online at their desk. The importance of wide reading of original papers should not be underestimated; relying solely on textbooks is not a good idea.

I have already noted that the MIPS software can be downloaded from www.wiley.com/go/mather4. That web site also contains compressed files of PowerPoint figures and tables which can be downloaded by lecturers and teachers. These files, one for each chapter, are compressed using the popular zip method. The four case studies included on the CD that accompanied the third edition of this book are also available on the web site as pdf files (with accompanying data sets).

I would like to acknowledge the help I have received from Rachel Ballard, Fiona Woods, Izzy Canning and Sarah Karim of John Wiley & Sons, Ltd. and Gayatri Shanker of Laserwords, who have looked after me during the production of this book with great patience and forbearance. Special thanks go to Alison Woodhouse, who copy-edited the manuscript in a masterly way. I must also thank my wife, who has been a great support over the past few years while the ‘book project’ was developing and before. My son James helped enormously in converting diagrams to colour. There are many other people who have had an input into the writing and production of this book, ranging from librarians to graphics designers and copy editors. I thank all of them for their assistance. Finally, a note to the reader: remote sensing is a captivating and intellectually demanding subject, but it is also fun. I have enjoyed teaching and researching in remote sensing for the best part of 30 years and I hope that my enjoyment of and enthusiasm for the subject comes across in the text and maybe inspires you to take a keen interest in whatever aspect of the subject that you are studying.

Needless to say, I take full responsibility for any errors that may be found in the book. I would be grateful if readers could advise me by email of any such errors or misrepresentations. My current email address can be found in the Contacts link on the book’s web page.

Paul M. Mather

Nottingham, January, 2010.

In the last decade, technological advances have led to a seemingly overwhelming amount of Earth observation information acquired at increasingly finer spatial, spectral and temporal scales. At the same time data processing and assimilation systems are continuously offering new and innovative ways of capturing, processing and delivering almost in real time this steady stream of geospatial data to end users, who may range from expert users to the untrained person requiring higher-level or value-added image products for decision making and problem solving.

This trend in geospatial technology development has naturally closed or at least narrowed the gap between remote sensing and GIS systems. RS data can provide both image maps for background display and estimates of the characteristics or values of environmental variables. These products can most profitably be used in conjunction with other environmental data such as rainfall or geology in the context of GIS.

I am honoured to contribute a new chapter, Chapter 10, on remote sensing and GIS integration to this new edition and would like to thank Prof. Paul Mather for giving me the opportunity to do so, and for guiding me through the process of ‘co-writing a book’. The new chapter draws from my experience in applying remote sensing and GIS techniques over a number of years while working on numerous research projects mainly conducted in Boston University’s Center for Remote Sensing. In this respect I am very grateful to Prof. Farouk El-Baz, director of the Center, who has been my adviser and colleague for the past two decades. I would also like to acknowledge the contributions and revisions of my colleagues Dr Francisco Estrada-Belli (Boston University), Dr Michael DiBlasi (Boston University), Dr Luisa Sernicola (University of Naples ‘L’Orientale’) and my PhD student Mr Ahmed Gaber (Tohoku University) in the case studies included in the new chapter 10. I hope that the combination of theoretical treatment and practical application of geospatial techniques will inspire the reader to create remote sensing and GIS-based solutions to a variety of environmental problems.

Magaly Koch

Boston, January 2010

List of Examples

Example 2.1Along-Track Scanning Radiometer (ATSR)Example 2.2MODISExample 2.3Landsat MSSExample 2.4Landsat ETM+Example 2.5SPOT-5 SupermodeExample 3.1Reading a Landsat ETM+ Image from CDExample 4.1Destriping – Linear MethodExample 4.2Orbital Geometry ModelExample 4.3Visualizing Polynomial SurfacesExample 5.1Linear Contrast StretchExample 6.1The Fourier TransformExample 7.1Moving Average FiltersExample 7.2Frequency Domain FilteringExample 8.1ISODATA Unsupervised ClassificationExample 8.2Supervised ClassificationExample 9.1DAIS 7915 Imaging Spectrometer DataExample 9.2Savitzky-Golay SmoothingExample 9.3Image Denoising Using the Wavelet TransformExample 9.4Lidar First/Last Return

‘I hope that posterity will judge me kindly, not only as to the things which I have explained but also as to those which I have intentionally omitted so as to leave to others the pleasure of discovery.’

René Descartes (1596-1660)

‘I am none the wiser, but I am much better informed.’

Queen Victoria (1819-1901)(After being addressed by Lord Rutherford on the state of modern physics)

‘Education is an admirable thing, but it is well to remember from time to time that nothing that is worth knowing can be taught.’

Oscar Wilde (1856-1900)

Web Page

A web page is associated with this book. Follow this link:

www.wiley.com/go/mather4 and you will find:

1. The MIPS image processing system software, including some test images,

2. PowerPoint Slide Shows containing all the figures and tables from each chapter,

3. A set of four example applications of remote sensing and GIS, including datasets.