Welcome to Scribd. Sign in or start your free trial to enjoy unlimited e-books, audiobooks & documents.Find out more
Buy Now $145.00
Standard view
Full view
of .
Look up keyword or section
Like this
1Activity

Table Of Contents

1.1 Introduction
1.1.1 Challenges in remote sensing
1.1.2 General concepts of machine learning
1.1.3 Paradigms in remote sensing
1.2 Supervised classification: algorithms and applications
1.2.1 Bayesian classification strategy
1.2.2 Neural networks
1.2.3 Support Vector Machines (SVM)
1.2.4 Use of multiple classifiers
1.3 Conclusion
References
2.1 Introduction
2.2 Kernels
2.2.1 Measuring similarity with kernels
2.2.2 Positive definite kernels
2.2.3 Constructing the reproducing kernel Hilbert space
2.2.4 Operations in RKHS
2.2.5 Kernel construction
2.2.6 Examples of kernels
2.3 The representer theorem
2.4 Learning with kernels
2.4.1 Support vector classification
2.4.2 Support vector regression
2.4.3 Gaussian processes
2.4.4 Multiple kernel learning
2.4.5 Structured prediction using kernels
2.4.6 Kernel principal component analysis
2.4.7 Applications of support vector algorithms
2.4.8 Available software
2.5 Conclusion
Supervised image classification
3.1 Introduction
3.2 Aspects of hyperspectral data and its acquisition
3.3 Hyperspectral remote sensing and supervised classification
3.4 Mathematical foundations of supervised classification
3.4.1 Empirical risk minimization
3.4.2 General bounds for a new risk minimization principle
3.4.3 Structural risk minimization
3.5 From structural risk minimization to a support vector machine algorithm
3.5.1 SRM for hyperplane binary classifiers
3.5.2 SVM algorithm
3.5.3 Kernel method
3.5.4 Hyperparameters
3.5.5 A toy example
3.5.6 Multi-class classifiers
3.5.7 Data centring
3.6 Benchmark hyperspectral data sets
3.6.1 The 4 class subset scene
3.6.2 The 16 class scene
3.6.3 The 9 class scene
3.7 Results
3.7.1 SVM implementation
3.7.2 Effect of hyperparameter d
3.7.3 Measure of accuracy of results
3.7.4 Classifier results for the 4 class subset scene and the 16 class full scene
3.7.6 Effect of training set size
3.7.7 Effect of simulated noisy data
3.8 Using spatial coherence
3.9 Why do SVMs perform better than other methods?
3.10 Conclusions
On training and evaluation of SVM for remote sensing applications
4.1 Introduction
4.2 Classification for thematic mapping
4.3 Overview of classification by a SVM
4.4.2 Training a SVM
4.4.3 Summary on training
4.5 Testing stage
4.5.1 General issues in testing
4.5.2 Specific issues for SVM classification
4.6 Conclusion
Kernel Fisher’s Discriminant with heterogeneous kernels
5.1 Introduction
5.2 Linear Fisher’s Discriminant
5.3 Kernel Fisher Discriminant
5.3.1 Mathematical programming formulation
5.4 Kernel Fisher’s Discriminant with heterogeneous kernels
5.5 Automatic kernel selection KFD algorithm
Algorithm 5.5.1 Automatic kernel selection KFD Algorithm (AKFD)
5.6 Numerical results
5.6.1 Dataset used: Purdue Campus data
5.6.2 Classifier design
5.6.3 Analysis of the results
5.7 Conclusion
Multi-temporal image classification with kernels
6.1 Introduction
6.1.1 Multi-temporal classification methods
6.1.2 Change detection methods
6.1.3 The proposed kernel-based framework
6.2 Multi-temporal classification and change detection with kernels
6.2.1 Problem statement and notation
6.2.2 Mercer’s kernels properties
6.2.3 Composite kernels for multi-temporal classification
6.2.4 Composite kernels for change detection
6.3 Contextual and multi-source data fusion with kernels
6.3.1 Composite kernels for integrating contextual information
6.3.2 Composite kernels for dealing with multi-source data
6.3.3 Remarks
6.4 Multi-temporal/-source urban monitoring
6.4.1 Model development and free parameter selection
6.4.2 Data collection and feature extraction
6.4.3 Multi-temporal image classification
6.4.4 Change detection
6.4.5 Classification maps
6.5 Conclusions
Target detection with kernels
7.1 Introduction
7.2 Kernel learning theory
7.3 Linear subspace-based anomaly detectors and their kernel versions
7.3.1 Principal component analysis
7.3.2 Kernel PCA subspace-based anomaly detection
7.3.3 Fisher linear discriminant analysis
7.3.4 Kernel fisher discriminant analysis
7.3.5 Eigenspace separation transform
7.3.6 Kernel eigenspace separation transform
7.3.7 RX algorithm
7.3.8 Kernel RX algorithm
7.4 Results
7.4.1 Simulated toy data
7.4.2 Hyperspectral imagery
7.5 Conclusion
One-class SVMs for hyperspectral anomaly detection
8.1 Introduction
8.2 Deriving the SVDD
8.2.1 The linear SVDD
8.2.2 The kernel-based SVDD
8.3 SVDD function optimization
8.4 SVDD algorithms for hyperspectral anomaly detection
8.4.1 Outline of algorithms
8.4.2 Dimensions for the background window
8.4.3 Estimating sigma
8.4.4 Normalized SVDD test statistic
8.5 Experimental results
8.6 Conclusions
Semi-supervised image classification
9.1 Introduction
9.2 Literature survey
9.2.2 Domain adaptation: partially-unsupervised methods
9.3 Proposed domain adaptation SVM
9.3.1 DASVM: problem definition and assumptions
9.3.2 DASVM: formulation
9.4 Proposed circular validation strategy
9.4.1 Circular validation strategy: rationale
9.4.2 Circular validation strategy: formulation
9.5 Experimental results
9.6 Discussions and conclusion
Mean kernels for semi-supervised remote sensing image classification
10.1 Introduction
10.2 Semi-supervised classification with mean kernels
10.2.1 Learning from labelled samples
10.2.2 Image clustering
10.2.3 Cluster similarity and the mean map
10.2.4 Composite sample-cluster kernels
10.2.5 Sample selection bias and the soft mean map
10.2.6 Summary of composite mean kernel methods
10.3 Experimental results
10.3.1 Model development
10.3.2 Results on synthetic data
10.3.3 Results on real data
Function approximation and regression
Kernel methods for unmixing hyperspectral imagery
11.1 Introduction
11.2 Mixing models
11.2.1 Areal mixtures
11.2.2 Intimate mixtures
11.3 Proposed kernel unmixing algorithm
11.3.1 Support vector data description for endmember extraction
11.3.2 Rate-distortion theory
11.4.2 AVIRIS data results
11.4.3 Processing times
11.5 Development of physics-based kernels for unmixing
11.5.1 Simplification of the albedo to reflectance transform
11.5.2 Kernel approximation of intimate mixtures
11.6 Physics-based kernel results
11.7 Summary
Kernel-based quantitative remote sensing inversion
12.1 Introduction
12.2 Typical kernel-based remote sensing inverse problems
12.2.1 Aerosol inverse problems
12.2.2 Land surface parameter retrieval problem
12.3 Well-posedness and ill-posedness
12.4 Regularization
12.4.1 Imposing a priori constraints on the solution
12.4.2 Tikhonov variational regularization
12.4.3 Direct regularization
12.4.4 Statistical regularization
12.5 Optimization techniques
12.5.1 Sparse inversion in l1 space
12.5.2 Optimization methods for l2 minimization model
12.6 Kernel-based BRDF model inversion
12.6.1 Inversion by NTSVD
12.6.2 Tikhonov regularized solution
12.6.3 Land surface parameter retrieval results
12.7 Aerosol particle size distribution function retrieval
12.8 Conclusion
Land and sea surface temperature estimation by support vector regression
13.1 Introduction
13.2 Previous work
13.2.1 LST and SST estimation from satellite data
13.2.2 Parameter optimization and error modelling for SVR
13.3 Methodology
13.3.1 SVR for LST and SST estimation
13.3.2 Automatic parameter optimization for SVR
13.3.3 Pointwise statistical modelling the SVR error
13.4 Experimental results
13.4.1 Data sets and experimental set-up
13.4.2 Parameter-optimization results
13.4.3 Results on the estimation of regression-error variance
13.5 Conclusions
Kernel-based feature extraction
Kernel multivariate analysis in remote sensing feature extraction
14.1 Introduction
14.2 Multivariate analysis methods
14.2.1 Principal component analysis (PCA)
14.2.2 Partial least squares
14.2.3 Canonical correlation analysis
14.2.4 Orthonormalized partial least squares
14.3 Kernel multivariate analysis
14.3.1 Kernel PCA
14.3.2 Kernel PLS
14.3.3 Kernel CCA
14.3.4 Kernel OPLS
14.3.5 Some considerations about Kernel MVA methods
14.4 Sparse Kernel OPLS
14.5 Experiments: pixel-based hyperspectral image classification
14.5.1 Data set description and experimental setup
14.5.2 Results description
14.6 Conclusions
KPCA algorithm for hyperspectral target/anomaly detection
15.1 Introduction
15.2 Motivation
15.2.1 Feature extraction of hyperspectral images
15.2.2 Introducing KM for hyperspectral image processing
15.2.3 Hyperspectral images for numerical experiments
15.3 Kernel-based feature extraction in hyperspectral images
15.3.1 Principal component analysis
15.3.2 Kernel mapping
15.3.3 Kernel Principal Component Analysis (KPCA)
15.4 Kernel-based target detection in hyperspectral images
15.4.1 The concept of target detection
15.4.2 Invariant subpixel material detector
15.4.3 Kernel invariant subpixel detection
15.5.3 Selective KPCA Feature Extraction for Anomaly Detection
15.6 Conclusions
16.1 Introduction
16.2 Related feature extractions
16.2.1 Linear discriminant analysis
16.2.2 Generalized discriminant analysis
16.2.3 Nonparametric weighted feature extraction
16.2.4 Fuzzy linear feature extraction
16.3 Kernel-based NWFE and FLFE
16.3.1 Kernel-based NWFE
16.3.2 Kernel-based FLFE
16.4 Eigenvalue resolution with regularization
16.5 Experiments
16.5.1 Data sets
16.5.2 Experiment design
16.5.3 Experiment results
16.6 Comments and conclusions
Index
P. 1
Kernel Methods for Remote Sensing Data Analysis

Kernel Methods for Remote Sensing Data Analysis

Ratings:
(0)
|Views: 2,736|Likes:
Published by Wiley
Kernel methods have long been established as effective techniques in the framework of machine learning and pattern recognition, and have now become the standard approach to many remote sensing applications. With algorithms that combine statistics and geometry, kernel methods have proven successful  across many different domains related to the analysis of images of the Earth acquired from airborne and satellite sensors, including natural resource control, detection and monitoring of anthropic infrastructures (e.g. urban areas), agriculture inventorying, disaster prevention and damage assessment, and anomaly and target detection.

 

Presenting the theoretical foundations of kernel methods (KMs) relevant to the remote sensing domain, this book serves as a practical guide to the design and implementation of these methods. Five distinct parts present state-of-the-art research related to remote sensing based on the recent advances in kernel methods, analysing the related methodological and practical challenges:

Part I introduces the key concepts of machine learning for remote sensing, and the theoretical and practical foundations of kernel methods. Part II explores supervised image classification including Super Vector Machines (SVMs), kernel discriminant analysis, multi-temporal image classification, target detection with kernels, and Support Vector Data Description (SVDD) algorithms for anomaly detection. Part III looks at semi-supervised classification with transductive SVM approaches for hyperspectral image classification and kernel mean data classification. Part IV examines regression and model inversion, including the concept of a kernel unmixing algorithm for hyperspectral imagery, the theory and methods for quantitative remote sensing inverse problems with kernel-based equations, kernel-based BRDF (Bidirectional Reflectance Distribution Function), and temperature retrieval KMs.  Part V deals with kernel-based feature extraction and provides a review of the principles of several multivariate analysis methods and their kernel extensions.

This book is aimed at engineers, scientists and researchers involved in remote sensing data processing, and also those working within machine learning and pattern recognition.

Kernel methods have long been established as effective techniques in the framework of machine learning and pattern recognition, and have now become the standard approach to many remote sensing applications. With algorithms that combine statistics and geometry, kernel methods have proven successful  across many different domains related to the analysis of images of the Earth acquired from airborne and satellite sensors, including natural resource control, detection and monitoring of anthropic infrastructures (e.g. urban areas), agriculture inventorying, disaster prevention and damage assessment, and anomaly and target detection.

 

Presenting the theoretical foundations of kernel methods (KMs) relevant to the remote sensing domain, this book serves as a practical guide to the design and implementation of these methods. Five distinct parts present state-of-the-art research related to remote sensing based on the recent advances in kernel methods, analysing the related methodological and practical challenges:

Part I introduces the key concepts of machine learning for remote sensing, and the theoretical and practical foundations of kernel methods. Part II explores supervised image classification including Super Vector Machines (SVMs), kernel discriminant analysis, multi-temporal image classification, target detection with kernels, and Support Vector Data Description (SVDD) algorithms for anomaly detection. Part III looks at semi-supervised classification with transductive SVM approaches for hyperspectral image classification and kernel mean data classification. Part IV examines regression and model inversion, including the concept of a kernel unmixing algorithm for hyperspectral imagery, the theory and methods for quantitative remote sensing inverse problems with kernel-based equations, kernel-based BRDF (Bidirectional Reflectance Distribution Function), and temperature retrieval KMs.  Part V deals with kernel-based feature extraction and provides a review of the principles of several multivariate analysis methods and their kernel extensions.

This book is aimed at engineers, scientists and researchers involved in remote sensing data processing, and also those working within machine learning and pattern recognition.

More info:

Publish date: Sep 3, 2009
Added to Scribd: Jun 04, 2013
Copyright:Traditional Copyright: All rights reservedISBN:9780470749005
List Price: $145.00 Buy Now

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
This book can be read on up to 6 mobile devices.
Buy the full version from:Amazon
See more
See less

03/02/2015

444

9780470749005

$145.00

USD

You're Reading a Free Preview
Pages 9 to 120 are not shown in this preview.
You're Reading a Free Preview
Pages 129 to 272 are not shown in this preview.
You're Reading a Free Preview
Pages 281 to 287 are not shown in this preview.
You're Reading a Free Preview
Pages 296 to 395 are not shown in this preview.
You're Reading a Free Preview
Pages 404 to 444 are not shown in this preview.

You're Reading a Free Preview

Download
scribd