Remote Sensing Digital Image Analysis

1 - Sources and Characteristics of Remote Sensing Image Data.- 1.1 Introduction to Data Sources.- 1.1.1 Characteristics of Digital Image Data.- 1.1.2 Spectral Ranges Commonly Used in Remote Sensing.- 1.1.3 Concluding Remarks.- 1.2 Weather Satellite Sensors.- 1.2.1 Polar Orbiting and Geosynchronous Satellites.- 1.2.2 The NOAA AVHRR (Advanced Very High Resolution Radiometer).- 1.2.3 The Nimbus CZCS (Coastal Zone Colour Scanner).- 1.2.4 GMS VISSR (Visible and Infrared Spin Scan Radiometer).- 1.3 Earth Resource Satellite Sensors in the Visible and Infrared Regions.- 1.3.1 The Landsat System.- 1.3.2 The Landsat Instrument Complement.- 1.3.3 The Return Beam Vidicon(RBV).- 1.3.4 The Multispectral Scanner (MSS).- 1.3.5 The Thematic Mapper (TM).- 1.3.6 The SPOT High Resolution Visible (HRV) Imaging Instrument.- 1.3.7 The Skylab S 192 Multispectral Scanner.- 1.3.8 The Heat Capacity Mapping Radiometer (HCMR).- 1.3.9 Marine Observation Satellite (MOS).- 1.3.10 Indian Remote Sensing Satellite (IRS).- 1.4 Aircraft Scanners in the Visible and Infrared Regions.- 1.4.1 General Considerations.- 1.4.2 The Daedalus AADS 1240/1260 Multispectral Line Scanner.- 1.4.3 The Airborne Thematic Mapper (ATM).- 1.4.4 The Thermal Infrared Multispectral Scanner (TIMS).- 1.4.5 The MDA MEIS-II Linear Array Aircraft Scanner.- 1.4.6 Imaging Spectrometers.- 1.5 Image Data Sources in the Microwave Region.- 1.5.1 Side Looking Airborne Radar and Synthetic Aperture Radar.- 1.5.2 TheSeasatSAR.- 1.5.3 Shuttle Imaging Radar-A (SIR-A).- 1.5.4 Shuttle Imaging Radar-B(SIR-B).- 1.5.5 ERS-1.- 1.5.6 JERS-1.- 1.5.7 Radarsat.- 1.5.8 Aircraft Imaging Radar Systems.- 1.6 Spatial Data Sources in General.- 1.6.1 Types of Spatial Data.- 1.6.2 Data Formats.- 1.6.3 Geographic Information Systems (GIS).- 1.6.4 The Challenge toImage Processing and Analysis.- 1.7 A Comparison of Scales in Digital Image Data.- References for Chapter 1.- Problems.- 2 - Error Correction and Registration of Image Data.- 2.1 Sources of Radiometric Distortion.- 2.1.1 The Effect of the Atmosphere on Radiation.- 2.1.2 Atmospheric Effects on Remote Sensing Imagery.- 2.1.3 Instrumentation Errors.- 2.2 Correction of Radiometric Distortion.- 2.2.1 Detailed Correction of Atmospheric Effects.- 2.2.2 Bulk Correction of Atmospheric Effects.- 2.2.3 Correction of Instrumentation Errors.- 2.3 Sources of Geometric Distortion.- 2.3.1 Earth Rotation Effects.- 2.3.2 Panoramic Distortion.- 2.3.3 Earth Curvature.- 2.3.4 Scan Time Skew.- 2.3.5 Variations in Platform Altitude, Velocity and Attitude.- 2.3.6 Aspect Ratio Distortion.- 2.3.7 Sensor Scan Nonlinearities.- 2.4 Correction of Geometric Distortion.- 2.4.1 Use of Mapping Polynomials for Image Correction.- 2.4.1.1 Mapping Polynomials and Ground Control Points.- 2.4.1.2 Resampling.- 2.4.1.3 Interpolation.- 2.4.1.4 Choice of Control Points.- 2.4.1.5 Example of Registration to a Map Grid.- 2.4.2 Mathematical Modelling.- 2.4.2.1 Aspect Ratio Correction.- 2.4.2.2 Earth Rotation Skew Correction.- 2.4.2.3 Image Orientation to North-South.- 2.4.2.4 Correction of Panoramic Effects.- 2.4.2.5 Combining the Corrections.- 2.5 Image Registration.- 2.5.1 Georeferencing and Geocoding.- 2.5.2 Image to Image Registration.- 2.5.3 Sequential Similarity Detection Algorithm.- 2.5.4 Example of Image to Image Registration.- 2.6 Miscellaneous Image Geometry Operations.- 2.6.1 Image Rotation.- 2.6.2 Scale Changing and Zooming.- References for Chapter 2.- Problems.- 3 - The Interpretation of Digital Image Data.- 3.1 Two Approaches to Interpretation.- 3.2 Forms of Imagery for Photointerpretation.- 3.3Computer Processing for Photointerpretation.- 3.4 An Introduction to Quantitative Analysis - Classification.- 3.5 Multispectral Space and Spectral Classes.- 3.6 Quantitative Analysis by Pattern Recognition.- 3.6.1 Pixel Vectors and Labelling.- 3.6.2 Unsupervised Classification.- 3.6.3 Supervised Classification.- References for Chapter 3.- Problems.- 4 - Radiometric Enhancement Techniques.- 4.1 Introduction.- 4.1.1 Point Operations and Look Up Tables.- 4.1.2 Scalar and Vector Images.- 4.2 The Image Histogram.- 4.3 Contrast Modification in Image Data.- 4.3.1 Histogram Modification Rule.- 4.3.2 Linear Contrast Enhancement.- 4.3.3 Saturating Linear Contrast Enhancement.- 4.3.4 Automatic Contrast Enhancement.- 4.3.5 Logarithmic and Exponential Contrast Enhancement.- 4.3.6 Piecewise Linear Contrast Modification.- 4.4 Histogram Equalization.- 4.4.1 Use of the Cumulative Histogram.- 4.4.2 Anomalies in Histogram Equalization.- 4.5 Histogram Matching.- 4.5.1 Principle of Histogram Matching.- 4.5.2 Image to Image Contrast Matching.- 4.5.3 Matching to a Mathematical Reference.- 4.6 Density Slicing.- 4.6.1 Black and White Density Slicing.- 4.6.2 Colour Density Slicing and Pseudocolouring.- References for Chapter 4.- Problems.- 5 - Geometric Enhancement Using Image Domain Techniques.- 5.1 Neighbourhood Operations.- 5.2 Template Operators.- 5.3 Geometric Enhancement as a Convolution Operation.- 5.4 ImageDomain Versus Fourier Transformation Approaches.- 5.5 Image Smoothing (Low Pass Filtering).- 5.5.1 Mean Value Smoothing.- 5.5.2 Median Filtering.- 5.6 Edge Detection and Enhancement.- 5.6.1 Linear Edge Detecting Templates.- 5.6.2 Spatial Derivative Techniques.- 5.6.2.1 The Roberts Operator.- 5.6.2.2 The Sobel Operator.- 5.6.3 Thinning, Linking and Edge Responses.- 5.6.4 EdgeEnhancement by Subtractive Smoothing.- 5.7 Line Detection.- 5.7.1 Linear Line Detecting Templates.- 5.7.2 Non-linear and Semi-linear Line Detecting Templates.- 5.8 General Convolution Filtering.- 5.9 Shape Detection.- References for Chapter 5.- Problems.- 6 - Multispectral Transformations of Image Data.- 6.1 The Principal Components Transformation.- 6.1.1 The Mean Vector and Covariance Matrix.- 6.1.2 A Zero Correlation, Rotational Transform.- 6.1.3 An Example - Some Practical Considerations.- 6.1.4 The Effect of an Origin Shift.- 6.1.5 Application of Principal Components in Image Enhancement and Display.- 6.1.6 The Taylor Method of Contrast Enhancement.- 6.1.7 Other Applications of Principal Components Analysis.- 6.2 The Kauth-Thomas Tasseled Cap Transformation.- 6.3 Image Arithmetic, Band Ratios and Vegetation Indices.- References for Chapter 6.- Problems.- 7 - Fourier Transformation of Image Data.- 7.1 Introduction.- 7.2 Special Functions.- 7.2.1 The Complex Exponential Function.- 7.2.2 The Dirac Delta Function.- 7.2.2.1 Properties of the Delta Function.- 7.2.3 The Heaviside Step Function.- 7.3 Fourier Series.- 7.4 The Fourier Transform.- 7.5 Convolution.- 7.5.1 The Convolution Integral.- 7.5.2 Convolution with an Impulse.- 7.5.3 The Convolution Theorem.- 7.6 Sampling Theory.- 7.7 The Discrete Fourier Transform.- 7.7.1 The Discrete Spectrum.- 7.7.2 Discrete Fourier Transform Formulae.- 7.7.3 Properties of the Discrete Fourier Transform.- 7.7.4 Computation of the Discrete Fourier Transform.- 7.7.5 Development of the Fast Fourier Transform Algorithm.- 7.7.6 Computational Cost of the Fast Fourier Transform.- 7.7.7 Bit Shuffling and Storage Considerations.- 7.8 The Discrete Fourier Transform of an Image.- 7.8.1 Definition.- 7.8.2 Evaluation of the Two Dimensional,Discrete Fourier Transform.- 7.8.3 The Concept of Spatial Frequency.- 7.8.4 Image Filtering for Geometric Enhancement.- 7.8.5 Convolution in Two Dimensions.- 7.9 Concluding Remarks.- References for Chapter 7.- Problems.- Chapters 8-Supervised Classification Techniques.- I. Standard Classification Algorithms.- 8.1 Steps in Supervised Classification.- 8.2 Maximum Likelihood Classification.- 8.2.1 Bayes'Classification.- 8.2.2 The Maximum Likelihood Decision Rule.- 8.2.3 Multivariate Normal Class Models.- 8.2.4 Decision Surfaces.- 8.2.5 Thresholds.- 8.2.6 Number of Training Pixels Required for Each Class.- 8.2.7 A Simple Illustration.- 8.3 Minimum Distance Classification.- 8.3.1 The Case of Limited Training Data.- 8.3.2 The Discriminant Function.- 8.3.3 Degeneration of Maximum Likelihood to Minimum Distance Classification.- 8.3.4 Decision Surfaces.- 8.3.5 Thresholds.- 8.4 Parallelepiped Classification.- 8.5 Classification Time Comparison of the Classifiers.- 8.6 The Mahalanobis Classifier.- 8.7 Table Look Up Classification.- II. More Advanced Considerations.- 8.8 Context Classification.- 8.8.1 The Concept of Spatial Context.- 8.8.2 Context Classification by Image Pre-Processing.- 8.8.3 Post Classification Filtering.- 8.8.4 Probabilistic Label Relaxation.- 8.8.4.1 The Basic Algorithm.- 8.8.4.2 The Neighbourhood Function.- 8.8.4.3 Determining the Compatibility Coefficients.- 8.8.4.4 The Final Step - Stopping the Process.- 8.8.4.5 Examples.- 8.9 Classification of Mixed Image Data.- 8.9.1 The Stacked Vector Approach.- 8.9.2 Statistical Methods.- 8.9.3 The Theory of Evidence.- 8.9.3.1 The Concept of Evidential Mass.- 8.9.3.2 Combining Evidence - the Orthogonal Sum.- 8.9.3.3 Decision Rule.- 8.10 Classification Using Neural Networks.- 8.10.1 Linear Discrimination.-8.10.1.1 Concept of a Weight Vector.- 8.10.1.2 Testing Class Membership.- 8.10.1.3 Training.- 8.10.1.4 Setting the Correction Increment.- 8.10.1.5 Classification - The Threshold Logic Unit.- 8.10.1.6 Multicategory Classification.- 8.10.2 Networks of Classifiers - Solutions of Nonlinear Problems.- 8.10.3 The Neural Network Approach.- 8.10.3.1 The Processing Element.- 8.10.3.2 Training the Neural Network - Backpropagation.- 8.10.3.3 Choosing the Network Parameters.- 8.10.3.4 Examples.- References for Chapter 8.- Problems.- 9 - Clustering and Unsupervised Classification.- 9.1 Delineation of Spectral Classes.- 9.2 Similarity Metrics and Clustering Criteria.- 9.3 The Iterative Optimization (Migrating Means) Clustering Algorithm.- 9.3.1 The Basic Algorithm.- 9.3.2 Mergings and Deletions.- 9.3.3 Splitting Elongated Clusters.- 9.3.4 Choice of Initial Cluster Centres.- 9.3.5 Clustering Cost.- 9.4 Unsupervised Classification and Cluster Maps.- 9.5 A Clustering Example.- 9.6 A Single Pass Clustering Technique.- 9.6.1 Single Pass Algorithm.- 9.6.2 Advantages and Limitations.- 9.6.3 Strip Generation Parameter.- 9.6.4 Variations on the Single Pass Algorithm.- 9.6.5 An Example.- 9.7 Agglomerative Hierarchical Clustering.- 9.8 Clustering by Histogram Peak Selection.- References for Chapter 9.- Problems 243.- 10 - Feature Reduction.- 10.1 Feature Reduction and Separability.- 10.2 Separability Measures for Multivariate Normal Spectral Class Models.- 10.2.1 Distribution Overlaps.- 10.2.2 Divergence.- 10.2.2.1 A General Expression.- 10.2.2.2 Divergence of a Pair of Normal Distributions.- 10.2.2.3 Use of Divergence for Feature Selection.- 10.2.2.4 A Problem with Divergence.- 10.2.3 The Jeffries-Matusita(JM) Distance.- 10.2.3.1 Definiton.- 10.2.3.2 Comparison of Divergence and JMDistance.- 10.2.4 Transformed Divergence.- 10.2.4.1 Definition.- 10.2.4.2 Relation between Transformed Divergence and Probability of Correct Classification.- 10.2.4.3 Use of Transformed Divergence in Clustering.- 10.3 Separability Measures for Minimum Distance Classification.- 10.4 Feature Reduction by Data Transformation.- 10.4.1 Feature Reduction Using the Principal Components Transformation.- 10.4.2 Canonical Analysis as a Feature Selection Procedure.- 10.4.2.1 Within Class and Among Class Covariance Matrices.- 10.4.2.2 A Separability Measure.- 10.4.2.3 The Generalised Eigenvalue Equation.- 10.4.2.4 An Example.- 10.4.3 Arithmetic Transformations.- References for Chapter 10.- Problems.- 11 - Image Classification Methodologies.- 11.1 Introduction.- 11.2 Supervised Classification.- 11.2.1 Outline.- 11.2.2 Determination of Training Data.- 11.2.3 Feature Selection.- 11.2.4 Detecting Multimodal Distributions.- 11.2.5 Presentation of Results.- 11.2.6 Effect of Resampling on Classification.- 11.3 Unsupervised Classification.- 11.3.1 Outline, and Comparison with Supervised Methods.- 11.3.2 Feature Selection.- 11.4 A Hybrid Supervised/Unsupervised Methodology.- 11.4.1 The Essential Steps.- 11.4.2 Choice of the Clustering Regions.- 11.4.3 Rationalisation of the Number of Spectral Classes.- 11.5 Assessment of Classification Accuracy.- 11.6 Case Study 1: Irrigated Area Determination.- 11.6.1 Background.- 11.6.2 The CSIRO-ORSER Image Analysis Software.- 11.6.3 The Study Region.- 11.6.4 Clustering.- 11.6.5 Signature Generation.- 11.6.6 Classification and Results.- 11.6.7 Concluding Remarks.- 11.7 Case Study 2: Multitemporal Monitoring of Bush Fires.- 11.7.1 Background.- 11.7.2 Simple Illustration of the Technique.- 11.7.3 The Study Area.- 11.7.4 Registration.- 11.7.5 PrincipalComponents Transformation.- 11.7.6 Classification of Principal Components Imagery.- References for Chapter 11.- Problems.- 12 - Knowledge-Based Image Analysis.- 12.1 Introduction.- 12.2 Knowledge Processing: Emulating Photointerpretation.- 12.3 Fundamentals of a Knowledge-Based Image Analysis System.- 12.3.1 Structure.- 12.3.2 Representation of Knowledge: Rules.- 12.3.3 The Inference Mechanism.- 12.4 Handling Multisource and Multisensor Data.- 12.5 An Example.- 12.5.1 Background.- 12.5.2 Rules as Justifiers for a Labelling Proposition.- 12.5.3 Endorsement of a Labelling Proposition.- 12.5.4 Knowledge Base and Results.- References for Chapter 12.- Problems.- Appendix A - Satellite Altitudes and Periods.- Referencesfor Appendix A.- Appendix B - Binary Representation of Decimal Numbers.- Appendix C - Essential Results from Vector and Matrix Algebra.- C.1 Definition of a Vector and a Matrix.- C.2 Properties of Matrices.- C.3 Multiplication, Addition and Subtraction of Matrices.- C.4 The Eigenvalues and Eigenvectors of a Matrix.- C.5 Some Important Matrix, Vector Operations.- C.6 An Orthogonal Matrix - the Concept of Matrix Transpose.- C. 7 Diagonalisation of a Matrix.- References for Appendix C.- Appendix D - Some Fundamental Material from Probability and Statistics.- D.1 Conditional Probability.- D.2 The Normal Probability Distribution.- D.2.1 The Univariate Case.- D.2.2 The Multivariate Case.- References for Appendix D.- Appendix E - Penalty Function Derivation of the Maximum Likelihood Decision Rule.- E.1 Loss Functions and Conditional Average Loss.- E.2 A Particular Loss Function.- References for Appendix E.- Appendix F - Real Time Image Processing in Interactive Image Display Systems.- F.1 Introduction.- F.2 The Use of Look Up Tables.- F.3 PointProcessing: The Use of Pipelines.- F.4 The Use of Feedback and Feedback Processing.- F.5 Neighbourhood Operations.- F.6 Limitations with Display Sub-System Processing.- References for Appendix F.