book

Classification Methods for Remotely Sensed Data, 3rd Edition

Name: Classification Methods for Remotely Sensed Data, 3rd Edition
ISBN: 9781040099117

by Taskin Kavzoglu, Brandt Tso, Paul M. Mather

September 2024

Intermediate to advanced

444 pages

21h 25m

English

CRC Press

Read now

Unlock full access

Cover
Half Title
Title Page
Copyright Page
Dedication
Table of Contents
Preface to the Third Edition
Preface to the Second Edition
Preface to the First Edition
Acknowledgments

Authors
Chapter 1 Fundamentals of Remote Sensing
1.1 Introduction to Remote Sensing1.1.1 Atmospheric Interactions1.1.2 Reflectance Properties of Surface Materials1.1.3 Spatial, Spectral, and Radiometric Resolution1.1.4 Scale Issues in Remote Sensing1.2 Optical Remote Sensing Systems1.3 Atmospheric Correction1.3.1 Dark Object Subtraction1.3.2 Modeling Techniques1.3.2.1 Modeling the Atmospheric Effect1.3.2.2 Steps in Atmospheric Correction1.4 Correction for Topographic Effects1.5 Remote Sensing in the Microwave Region1.6 Radar Fundamentals1.6.1 SLAR Image Resolution1.6.2 Geometric Effects on Radar Images1.6.3 Factors Affecting Radar Backscatter1.6.3.1 Surface Roughness1.6.3.2 Surface Conductivity1.6.3.3 Parameters of the Radar Equation1.7 Imaging Radar Polarimetry1.7.1 Radar Polarization State1.7.2 Polarization Synthesis1.7.3 Polarization Signatures1.8 Radar Speckle Suppression1.8.1 Multilook Processing1.8.2 Filters for Speckle SuppressionReferences
Chapter 2 Pattern Recognition Principles
2.1 A Terminological Introduction2.2 Taxonomy of Classification Techniques2.3 Fundamental Pattern Recognition Techniques2.3.1 Unsupervised Methods2.3.1.1 The k-Means Algorithm2.3.1.2 Fuzzy C-Means Clustering2.3.2 Supervised Methods2.3.2.1 Parallelepiped Method2.3.2.2 Minimum Distance Classifier2.3.2.3 Maximum Likelihood Classifier2.3.2.4 Fuzzy Maximum Likelihood Classifier2.4 Spectral Unmixing2.5 Ensemble Classifiers2.6 Incorporation of Ancillary Information2.6.1 Use of Texture and Context2.6.2 Using Ancillary Multisource Data2.7 EpilogueReferences
Chapter 3 Dimensionality Reduction: Feature Extraction and Selection
3.1 Feature Extraction3.1.1 Principal Component Analysis3.1.2 Minimum/Maximum Autocorrelation Factors3.1.3 Maximum Noise Fraction (MNF) Transformation3.1.4 Independent Component Analysis3.1.5 Projection Pursuit3.2 Feature Selection3.2.1 Greedy Search Methods3.2.2 Simulated Annealing3.2.3 Separability Indices3.2.4 Filter-Based Methods3.2.4.1 Correlation-Based Feature Selection3.2.4.2 Information Gain3.2.4.3 Gini Impurity Index3.2.4.4 Minimum Redundancy-Maximum Relevance3.2.4.5 Chi-Square Test3.2.4.6 Relief-F3.2.4.7 Symmetric Uncertainty3.2.4.8 Fisher’s Test3.2.4.9 OneR3.2.5 Wrappers3.2.5.1 Genetic Algorithm3.2.5.2 Particle Swarm Optimization3.2.5.3 Feature Selection with SVMs3.2.6 Embedded Methods3.2.6.1 K-Nearest Neighbor-Based Feature Selection3.2.6.2 Feature Selection with Ensemble Learners3.2.6.3 Hilbert-Schmidt Independence Criterion with Lasso3.3 Concluding RemarksReferences
Chapter 4 Multisource Image Fusion and Classification
4.1 Image Fusion4.1.1 Image Fusion Methods4.1.1.1 PCA-Based Image Fusion4.1.1.2 IHS-Based Image Fusion4.1.1.3 Brovey Transform4.1.1.4 Gram-Schmidt Transform4.1.1.5 Wavelet Transform4.1.1.6 Deep Learning for Image Fusion4.1.2 Assessment of Fused Image Quality4.1.3 Performance Evaluation of Fusion Methods4.2 Multisource Classification Using the Stacked-Vector Method4.3 The Extension of Bayesian Classification Theory4.3.1 An Overview4.3.1.1 Feature Extraction4.3.1.2 Probability or Evidence Generation4.3.1.3 Multisource Consensus4.3.2 Bayesian Multisource Classification Mechanism4.3.3 A Refined Multisource Bayesian Model4.3.4 Multisource Classification Using the MRF4.3.5 Assumption of Inter-Source Independence4.4 Evidential Reasoning4.4.1 Concept Development4.4.2 Belief Function and Belief Interval4.4.3 Evidence Combination4.4.4 Decision Rules for Evidential Reasoning4.5 Dealing with Source Reliability4.5.1 Using Classification Accuracy4.5.2 Use of Class Separability4.5.3 Data Information Class Correspondence Matrix4.6 Concluding Remarks and Future TrendsReferences
Chapter 5 Support Vector Machines
5.1 Linear Classification5.1.1 The Separable Case5.1.2 The Nonseparable Case5.2 Nonlinear Classification and Kernel Functions5.2.1 Nonlinear SVMs5.2.2 Kernel Functions5.3 Parameter Determination5.3.1 t-Fold Cross-Validations5.3.2 Bound on Leave-One-Out Error5.3.3 Grid Search5.3.4 Gradient Descent Method5.4 Multiclass Classification5.4.1 One-against-One, One-against-Others, and DAG5.4.2 Multiclass SVMs5.4.2.1 Vapnik’s Approach5.4.2.2 Methodology of Crammer and Singer5.5 Relevance Vector Machines5.6 Twin Support Vector Machines5.7 Deep Support Vector Machines5.8 Concluding RemarksReferences
Chapter 6 Decision Trees
6.1 ID3, C4.5, and SEE5.0 Decision Trees6.1.1 ID36.1.2 C4.56.1.3 SEE5.0 (C5.0)6.2 CHAID6.3 CART6.4 QUEST6.4.1 Split Point Selection6.4.2 Attribute Selection6.5 Tree Induction from Artificial Neural Networks6.6 Pruning Decision Trees6.6.1 Reduced Error Pruning6.6.2 Pessimistic Error Pruning6.6.3 Error-Based Pruning6.6.4 Cost Complexity Pruning6.6.5 Minimal Error Pruning6.7 Ensemble Methods6.7.1 Boosting6.7.2 Random Forest6.7.3 Rotation Forest6.7.4 Canonical Correlation Forest6.7.5 Extreme Gradient Boosting6.7.6 Light Gradient Boosting Machines6.7.7 Gradient Boosting Machines6.7.8 Categorical Boosting6.7.9 Natural Gradient Boosting6.8 Concluding RemarksReferences
Chapter 7 Deep Learning
7.1 Fundamentals7.1.1 Stochastic Gradient Descent7.1.2 Backpropagation7.1.3 Regularization7.1.3.1 Weight Decay7.1.3.2 Dropout7.1.3.3 Data Augmentation7.1.3.4 Early Stopping7.1.4 Activation Functions7.1.5 Loss Functions7.2 Neural Network Architectures7.2.1 Multilayer Perceptron7.2.2 Convolutional Neural Networks7.2.2.1 Convolutional Layers7.2.2.2 Pooling Layers7.2.2.3 Fully Connected Layers7.2.2.4 Receptive Field and Feature Map7.2.2.5 Training CNNs7.2.2.6 Data Structures in CNNs7.2.2.7 Evolving Trends in CNN Design7.2.3 Recurrent Neural Networks7.2.3.1 Long- and Short-Term Memory7.2.3.2 Gated Recurrent Unit7.2.4 Vision Transformers7.2.5 Deep Multilayer Perceptron7.2.6 Generative Adversarial Networks7.2.7 Deep Autoencoders7.2.7.1 Undercomplete Autoencoders7.2.7.2 Regularized Autoencoders7.2.7.3 Sparse Autoencoders7.2.7.4 Denoising Autoencoders7.2.7.5 Variational Autoencoders7.3 Learning Paradigms7.3.1 Transfer Learning7.3.2 Semi-Supervised Learning7.3.3 Reinforcement Learning7.3.4 Active Learning7.3.5 Multitask Learning7.4 Application of DL in Remote Sensing7.4.1 Semantic Segmentation7.4.2 Object Detection7.4.3 Scene Classification7.4.4 Change Detection7.5 Concluding RemarksReferences
Chapter 8 Object-Based Image Analysis
8.1 Clustering-Based Segmentation8.1.1 Mean-Shift Algorithm8.1.2 Superpixel Segmentation8.2 Thresholding-Based Segmentation8.3 Edge-Based Segmentation8.4 Watershed Segmentation8.5 Region-Based Segmentation8.5.1 Region Splitting and Merging8.5.2 Region Growing8.5.3 Multiresolution Segmentation8.6 Hybrid Segmentation8.7 Evaluation of Segmentation Quality8.7.1 Supervised Approach8.7.2 Unsupervised Approach8.7.2.1 Estimation of the Scale Parameter8.7.2.2 Global Score8.7.2.3 Overall Goodness F-Measure8.8 Concluding RemarksReferences
Chapter 9 Hyperparameter Optimization
9.1 What Is Hyperparameter Optimization?9.2 Hyperparameter Optimization Techniques9.2.1 Model-Free Algorithms9.2.1.1 Trial-and-Error (Manual Testing)9.2.1.2 Grid Search9.2.1.3 Random Search9.2.2 Gradient-Based Optimization9.2.3 Bayesian Optimization9.2.4 Multifidelity Optimization9.2.4.1 Successive Halving9.2.4.2 Hyperband9.2.5 Metaheuristic Algorithms9.2.5.1 Genetic Algorithm9.2.5.2 Particle Swarm Optimization9.3 Challenges in Hyperparameter Optimization9.4 Concluding RemarksReferences
Chapter 10 Accuracy Assessment and Model Explainability
10.1 Accuracy Assessment10.1.1 Sampling Scheme and Spatial Autocorrelation10.1.2 Sample Size, Scale, and Spatial Variability10.1.3 Adequacy of Training and Testing Data10.1.4 Conventional Accuracy Analysis10.1.5 Accuracy Analysis for Machine Learning10.1.6 Fuzzy Accuracy Assessment10.1.7 Object-Based Accuracy Assessment10.2 Comparison of Thematic Maps10.2.1 McNemar’s Test10.2.2 z-Test10.2.3 Wilcoxon Signed-Ranks Test10.2.4 5 × 2-Cross-Validation t-Test10.2.5 Friedman Test10.3 Explainability Methods10.3.1 SHapley Additive exPlanations10.3.2 Partial Dependence Plot10.3.3 Pairwise Interaction Importance10.3.4 Permutation-Based Feature Importance10.3.5 Local Interpretable Model-Agnostic Explanations (LIME)10.4 A Case Study for Accuracy Assessment and XAI10.5 Conclusions and Guidelines for Best PracticeReferences
Index

Overview

The new edition of the bestselling Classification Methods for Remotely Sensed Data covers current state-of-the-art machine learning algorithms and developments in the analysis of remotely sensed data, and presents new AI-based analysis tools and metrics together with ongoing debates on accuracy assessment strategies and XAI methods.

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781040099117

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills