Techniques and Applications of Hyperspectral Image Analysis -- Contents -- Preface -- List of Contributors -- List of Abbreviations -- 1 Multivariate Images, Hyperspectral Imaging: Background and Equipment -- 1.1 Introduction -- 1.2 Digital Images, Multivariate Images and Hyperspectral Images -- 1.3 Hyperspectral Image Generation -- 1.3.1 Introduction -- 1.3.2 Point Scanning Imaging -- 1.3.3 Line Scanning Imaging -- 1.3.4 Focal Plane Scanning Imaging -- 1.4 Essentials of Image Analysis Connecting Scene and Variable Spaces -- References -- 2 Principles of Multivariate Image Analysis (MIA) in Remote Sensing, Technology and Industry -- 2.1 Introduction -- 2.1.1 MIA Approach: Synopsis -- 2.2 Dataset Presentation -- 2.2.1 Master Dataset: Rationale -- 2.2.2 Montmorency Forest, Quebec, Canada: Forestry Background -- 2.3 Tools in MIA -- 2.3.1 MIA Score Space Starting Point -- 2.3.2 Colour-slice Contouring in Score Cross-plots: a 3-D Histogram -- 2.3.3 Brushing: Relating Different Score Cross-plots -- 2.3.4 Joint Normal Distribution (or Not) -- 2.3.5 Local Models/Local Modelling: a Link to Classification -- 2.4 MIA Analysis Concept: Master Dataset Illustrations -- 2.4.1 A New Topographic Map Analogy -- 2.4.2 MIA Topographic Score Space Delineation of Single Classes -- 2.4.3 MIA Delineation of End-member Mixing Classes -- 2.4.4 Which to Use? When? How? -- 2.4.5 Scene-space Sampling in Score Space -- 2.5 Conclusions -- References -- 3 Clustering and Classification in Multispectral Imaging for Quality Inspection of Postharvest Products -- 3.1 Introduction to Multispectral Imaging in Agriculture -- 3.1.1 Measuring Quality -- 3.1.2 Spectral Imaging in Agriculture -- 3.2 Unsupervised Classification of Multispectral Images -- 3.2.1 Unsupervised Classification with FCM -- 3.2.2 FCM Clustering -- 3.2.3 cFCM Clustering -- 3.2.4 csiFCM.

3.2.5 Combining Spectral and Spatial Information -- 3.2.6 sgFCM Clustering -- 3.3 Supervised Classification of Multispectral Images -- 3.3.1 Multivariate Image Analysis for Training Set Selection -- 3.3.2 FEMOS -- 3.3.3 Experiment with a Multispectral Image of Pine and Spruce Wood -- 3.3.4 Clustering with FEMOS Procedure -- 3.4 Visualization and Coloring of Segmented Images and Graphs: Class Coloring -- 3.5 Conclusions -- References -- 4 Self-modeling Image Analysis with SIMPLISMA -- 4.1 Introduction -- 4.2 Materials and Methods -- 4.2.1 FTIR Microscopy -- 4.2.2 SIMS Imaging of a Mixture of Palmitic and Stearic Acids on Aluminum foil -- 4.2.3 Data Analysis -- 4.3 Theory -- 4.4 Results and Discussion -- 4.4.1 FTIR Microscopy Transmission Data of a Polymer Laminate -- 4.4.2 FTIR Reflectance Data of a Mixture of Aspirin and Sugar -- 4.4.3 SIMS Imaging of a Mixture of Palmitic and Stearic Acids on Aluminum Foil -- 4.5 Conclusions -- References -- 5 Multivariate Analysis of Spectral Images Composed of Count Data -- 5.1 Introduction -- 5.2 Example Datasets and Simulations -- 5.3 Component Analysis -- 5.4 Orthogonal Matrix Factorization -- 5.4.1 PCA and Related Methods -- 5.4.2 PCA of Arbitrary Factor Models -- 5.4.3 Maximum Likelihood PCA (MLPCA) -- 5.4.4 Weighted PCA (WPCA) -- 5.4.5 Principal Factor Analysis (PFA) -- 5.4.6 Selecting the Number of Components -- 5.5 Maximum Likelihood Based Approaches -- 5.5.1 Poisson Non-negative Matrix Factorization (PNNMF) -- 5.5.2 Iteratively Weighted Least Squares (IWLS) -- 5.5.3 NNMF: Gaussian Case (Approximate Noise) -- 5.5.4 Factored NNMF: Gaussian Case (Approximate Data) -- 5.5.5 Alternating Least Squares (ALS) -- 5.5.6 Performance Comparisons -- 5.6 Conclusions -- Acknowledgements -- References -- 6 Hyperspectral Image Data Conditioning and Regression Analysis -- 6.1 Introduction.

6.2 Terminology 1286.3 Multivariate Image -- 6.3 Multivariate Image Regression -- 6.3.1 Regression Diagnostics -- 6.3.2 Differences between Normal Calibration and Image Calibration -- 6.4 Data Conditioning -- 6.4.1 Reflectance Transformation and Standardization -- 6.4.2 Spectral Transformations -- 6.4.3 Data Clean-up -- 6.4.4 Data Conditioning Summary -- 6.5 PLS Regression Optimization -- 6.5.1 Data Subset Selection -- 6.5.2 Pseudorank Determination -- 6.6 Regression Examples -- 6.6.1 Artificial Ternary Mixture -- 6.6.2 Commercial Cheese Samples -- 6.6.3 Wheat Straw Wax -- 6.7 Conclusions -- Acknowledgements -- References -- 7 Principles of Image Cross-validation (ICV): Representative Segmentation of Image Data Structures -- 7.1 Introduction -- 7.2 Validation Issues -- 7.2.1 2-way -- 7.2.2 MIA/MIR -- 7.3 Case Studies -- 7.3.1 Case 1: Full Y-image -- 7.3.2 Case 2: Critical Segmentation Issues -- 7.3.3 Case 3: Y-composite -- 7.3.4 Case 4: Image Data Structure Sampling -- 7.4 Discussion and Conclusions -- 7.5 Reflections on 2-way Cross-validation -- References -- 8 Detection, Classification, and Quantification in Hyperspectral Images Using Classical Least Squares Models -- 8.1 Introduction -- 8.2 CLS Models -- 8.2.1 CLS -- 8.2.2 ELS -- 8.2.3 GLS -- 8.3 Detection, Classification, and Quantification -- 8.3.1 Detection -- 8.3.2 Classification -- 8.3.3 Quantification -- 8.4 Conclusions -- Acknowledgements -- References -- 9 Calibration Standards and Image Calibration -- 9.1 Introduction -- 9.2 The Need for Calibration in General -- 9.3 The Need for Image Calibration -- 9.4 Resolution in Hyperspectral Images -- 9.5 Spectroscopic Definitions -- 9.6 Calibration Standards -- 9.7 Calibration in Hyperspectral Images -- 9.8 Conclusions -- References -- 10 Multivariate Movies and their Applications in Pharmaceutical and Polymer Dissolution Studies.

10.1 Introduction -- 10.1.1 Introducing the Time Axis -- 10.1.2 Data Structure and Reduction -- 10.1.3 Compression of Spectra -- 10.1.4 Space Dimensions -- 10.1.5 Time Dimension -- 10.1.6 Simultaneous Compression of all Variables -- 10.2 Applications: Solvent Diffusion and Pharmaceutical Studies -- 10.2.1 Solvent Diffusion in Polymers -- 10.2.2 Optical and NMR Studies -- 10.2.3 Line Imaging -- 10.2.4 Global MIR Imaging Studies of Solvent Intake -- 10.3 Drug Release -- 10.3.1 ATR-FTIR Imaging -- 10.4 Conclusions -- Acknowledgement -- References -- 11 Multivariate Image Analysis of Magnetic Resonance Images: Component Resolution with the Direct Exponential Curve Resolution Algorithm (DECRA) -- 11.1 Introduction -- 11.2 DECRA Approach -- 11.3 DECRA Algorithm -- 11.4 1H Relaxation -- 11.5 T1 Transformation -- 11.6 Imaging Methods -- 11.7 Phantom Images -- 11.7.1 T2 Series -- 11.7.2 T1 Series -- 11.8 Brain Images -- 11.8.1 T2 Series -- 11.8.2 T1 Series -- 11.9 Regression Analysis -- 11.10 Conclusions -- References -- 12 Hyperspectral Imaging Techniques: an Attractive Solution for the Analysis of Biological and Agricultural Materials -- 12.1 Introduction -- 12.2 Sample Characterization and Chemical Species Distribution -- 12.2.1 Analysis of Fruit -- 12.2.2 Analysis of Kernels -- 12.2.3 Analysis of Food and Feed Mixtures -- 12.3 Detecting Contamination and Defects in Agro-food Products -- 12.3.1 Detecting Contamination in Meat Products -- 12.3.2 Detecting Contamination and Defects in Fruit -- 12.3.3 Detecting Contamination and Defects in Cereals -- 12.3.4 Detecting Contamination in Compound Feed -- 12.4 Other Agronomic and Biological Applications -- 12.5 Conclusion -- References.

13 Application of Multivariate Image Analysis in Nuclear Medicine: Principal Component Analysis (PCA) on Dynamic Human Brain Studies with Positron Emission Tomography (PET) for Discrimination of Areas of Disease at High Noise Levels -- 13.1 Introduction -- 13.2 PET -- 13.2.1 History -- 13.2.2 Principles -- 13.2.3 Scanning Modes in PET -- 13.2.4 Analysis of PET Data/Images -- 13.3 PCA -- 13.3.1 History -- 13.3.2 Definition -- 13.3.3 Pre-processing and Scaling -- 13.3.4 Noise Pre-normalization -- 13.4 Application of PCA in PET -- 13.4.1 SWPCA -- 13.4.2 VWPCA -- 13.4.3 MVWPCA -- 13.5 Conclusions -- References -- 14 Near Infrared Chemical Imaging: Beyond the Pictures -- 14.1 Introduction -- 14.2 Data Measurement -- 14.3 Selection of Samples and Acquisition Schemes -- 14.4 What, How Much and Where -- 14.5 Data Analysis and the Underlying Structure of the Data -- 14.6 Imaging with Statistically Meaningful Spatial Dimensions -- 14.7 Chemical Contrast -- 14.8 Measure, Count and Compare -- 14.9 Correlating Data to Sample Performance and/or Behavior: the Value of NIRCI Data -- 14.10 Conclusions -- References -- Index -- Plates.