Hyperspectral Data Processing_ Algorithm Design and Analysis.pdf-资料库

cgs425449830-10636420-4744300845401230845.pdf-第1页.png

第1页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第2页.png

第2页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第3页.png

第3页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第4页.png

第4页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第5页.png

第5页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第6页.png

第6页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第7页.png

第7页 / 共1151页

cgs425449830-10636420-4744300845401230845.pdf-第8页.png

第8页 / 共1151页

HYPERSPECTRAL DATA PROCESSING

HYPERSPECTRAL DATA PROCESSING Algorithm Design and Analysis Chein-I Chang University of Maryland, Baltimore County (UMBC), Maryland, USA

Copyright # 2013 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, NJ Published simultaneously in Canada 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, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and speciﬁcally disclaim any implied warranties of merchantability or ﬁtness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of proﬁt or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Chang, Chein-I. Hyperspectral data processing : algorithm design and analysis / Chein-I Chang. p. cm. Includes bibliographical references and index. ISBN 978-0-471-69056-6 (hardback) 1. Image processing–Digital techniques. 2. Spectroscopic imaging. 3. Signal processing. I. Chang, Chein-I. Hyperspectral imaging. II. Title. TA1637.C4776 2012 621.39’94–dc23 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 2011043896

This book is dedicated to members of my family, speciﬁcally my mother who provided me with her timeless support and encouragement during the course of preparing this book. It is also dedicated to all of my students who have contributed to this book.

Contents PREFACE xxiii 1 OVERVIEW AND INTRODUCTION 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Overview Issues of Multispectral and Hyperspectral Imageries Divergence of Hyperspectral Imagery from Multispectral Imagery 1.3.1 Misconception: Hyperspectral Imaging is a Natural Extension of Multispectral Imaging Pigeon-Hole Principle: Natural Interpretation of Hyperspectral Imaging 1.3.2 Scope of This Book Book’s Organization 1.5.1 1.5.2 1.5.3 1.5.4 1.5.5 1.5.6 1.5.7 1.5.8 Part I: Preliminaries Part II: Endmember Extraction Part III: Supervised Linear Hyperspectral Mixture Analysis Part IV: Unsupervised Hyperspectral Analysis Part V: Hyperspectral Information Compression Part VI: Hyperspectral Signal Coding Part VII: Hyperspectral Signal Feature Characterization Applications 1.5.8.1 1.5.8.2 Chapter 30: Applications of Target Detection Chapter 31: Nonlinear Dimensionality Expansion to Multispectral Imagery Chapter 32: Multispectral Magnetic Resonance Imaging 1.5.8.3 Laboratory Data to be Used in This Book 1.6.1 Laboratory Data 1.6.2 Cuprite Data 1.6.3 NIST/EPA Gas-Phase Infrared Database Real Hyperspectral Images to be Used in this Book 1.7.1 AVIRIS Data 1.7.1.1 1.7.1.2 HYDICE Data Cuprite Data Purdue’s Indiana Indian Pine Test Site 1.7.2 Notations and Terminologies to be Used in this Book 1 2 3 4 4 5 7 10 10 12 13 13 15 16 17 17 17 18 19 19 19 19 19 20 20 21 25 26 29 vii

viii I: PRELIMINARIES 2 FUNDAMENTALS OF SUBSAMPLE AND MIXED SAMPLE ANALYSES Contents 31 33 33 35 35 38 39 41 43 45 45 46 48 54 54 56 57 57 58 59 60 63 63 65 67 69 72 72 73 75 77 78 79 79 80 83 84 87 91 95 99 2.1 2.2 Introduction Subsample Analysis 2.2.1 2.2.2 2.3 Mixed Sample Analysis Pure-Sample Target Detection Subsample Target Detection 2.2.2.1 2.2.2.2 Subsample Target Detection: Constrained Energy Minimization (CEM) Adaptive Matched Detector (AMD) Adaptive Subspace Detector (ASD) 2.2.3 2.3.1 2.3.2 Fisher’s Linear Discriminant Analysis (FLDA) Support Vector Machines (SVM) Classiﬁcation with Hard Decisions 2.3.1.1 2.3.1.2 Classiﬁcation with Soft Decisions 2.3.2.1 2.3.2.2 Orthogonal Subspace Projection (OSP) Target-Constrained Interference-Minimized Filter (TCIMF) 2.4 2.5 Kernel-Based Classiﬁcation 2.4.1 2.4.2 2.4.3 Conclusions Kernel Trick Used in Kernel-Based Methods Kernel-Based Fisher’s Linear Discriminant Analysis (KFLDA) Kernel Support Vector Machine (K-SVM) 3 THREE-DIMENSIONAL RECEIVER OPERATING CHARACTERISTICS (3D ROC) ANALYSIS 3.1 3.2 3.3 3.4 3.5 3.6 3.7 Introduction Neyman–Pearson Detection Problem Formulation ROC Analysis 3D ROC Analysis Real Data-Based ROC Analysis 3.5.1 3.5.2 3.5.3 3.5.4 How to Generate ROC Curves from Real Data How to Generate Gaussian-Fitted ROC Curves How to Generate 3D ROC Curves How to Generate 3D ROC Curves for Multiple Signal Detection and Classiﬁcation Examples 3.6.1 3.6.2 3.6.3 3.6.4 Conclusions Hyperspectral Target Detection Linear Hyperspectral Mixture Analysis Hyperspectral Imaging 3.6.1.1 3.6.1.2 Magnetic Resonance (MR) Breast Imaging Breast Tumor Detection 3.6.2.1 3.6.2.2 Brain Tissue Classiﬁcation Chemical/Biological Agent Detection Biometric Recognition

Contents 4 DESIGN OF SYNTHETIC IMAGE EXPERIMENTS 4.1 4.2 4.3 4.4 Introduction Simulation of Targets of Interest 4.2.1 4.2.2 Six Scenarios of Synthetic Images 4.3.1 4.3.2 Simulation of Synthetic Subsample Targets Simulation of Synthetic Mixed-Sample Targets Panel Simulations Three Scenarios for Target Implantation (TI) 4.3.2.1 4.3.2.2 4.3.2.3 Scenario TI1 (Clean Panels Implanted into Clean Background) Scenario TI2 (Clean Panels Implanted into Noisy Background) Scenario TI3 (Gaussian Noise Added to Clean Panels Implanted into Clean Background) 4.3.3 Three Scenarios for Target Embeddedness (TE) 4.3.3.1 4.3.3.2 4.3.3.3 Scenario TE1 (Clean Panels Embedded in Clean Background) Scenario TE2 (Clean Panels Embedded in Noisy Background) Scenario TE3 (Gaussian Noise Added to Clean Panels Embedded in Background) Applications 4.4.1 4.4.2 4.4.3 Endmember Extraction Linear Spectral Mixture Analysis (LSMA) 4.4.2.1 4.4.2.2 Target Detection 4.4.3.1 4.4.3.2 Mixed Pixel Classiﬁcation Mixed Pixel Quantiﬁcation Subpixel Target Detection Anomaly Detection 4.5 Conclusions 5 VIRTUAL DIMENSIONALITY OF HYPERSPECTRAL DATA 5.1 5.2 5.3 Introduction Reinterpretation of VD VD Determined by Data Characterization-Driven Criteria 5.3.1 Eigenvalue Distribution-Based Criteria 5.3.1.1 5.3.1.2 Thresholding Energy Percentage Thresholding Difference between Normalized Correlation Eigenvalues and Normalized Covariance Eigenvalues Finding First Sudden Drop in the Normalized Eigenvalue Distribution 5.3.1.3 5.3.2 5.3.3 5.3.4 5.3.5 Singular Value Decomposition (SVD) Principal Components Analysis (PCA) Eigen-Based Component Analysis Criteria 5.3.2.1 5.3.2.2 Factor Analysis: Malinowski’s Error Theory Information Theoretic Criteria (ITC) 5.3.4.1 5.3.4.2 Gershgorin Radius-Based Methods 5.3.5.1 5.3.5.2 AIC MDL Thresholding Gershgorin Radii Thresholding Difference Gershgorin Radii between RLL and KLL ix 101 102 103 103 104 104 104 106 106 107 108 108 109 109 110 112 112 113 114 114 114 114 122 123 124 124 126 126 127 127 128 128 128 128 129 129 130 131 131 131 134 134

x Contents HFC Method Discussions on Data Characterization-Driven Criteria 5.3.6 5.3.7 VD Determined by Data Representation-Driven Criteria 5.4.1 5.4.2 5.4.3 Synthetic Image Experiments 5.5.1 Orthogonal Subspace Projection (OSP) Signal Subspace Estimation (SSE) Discussions on OSP and SSE/HySime Data Characterization-Driven Criteria 5.5.1.1 5.5.1.2 5.5.2 Data Representation-Driven Criteria VD Estimated for Real Hyperspectral Images Conclusions Target Implantation (TI) Scenarios Target Embeddedness (TE) Scenarios 5.4 5.5 5.6 5.7 6 DATA DIMENSIONALITY REDUCTION 6.1 6.2 Introduction Dimensionality Reduction by Second-Order Statistics-Based Component Analysis Transforms 6.2.1 6.2.2 Principal Components Analysis Standardized Principal Components Analysis Singular Value Decomposition Eigen Component Analysis Transforms 6.2.1.1 6.2.1.2 6.2.1.3 Signal-to-Noise Ratio-Based Components Analysis Transforms 6.2.2.1 6.2.2.2 Maximum Noise Fraction Transform Noise-Adjusted Principal Component Transform 6.3 6.4 6.5 6.6 6.7 Statistics-Prioritized ICA-DR (SPICA-DR) Random ICA-DR Initialization Driven ICA-DR Sphering Third-Order Statistics-Based Skewness Fourth-Order Statistics-Based Kurtosis High-Order Statistics Algorithm for Finding Projection Vectors Dimensionality Reduction by High-Order Statistics-Based Components Analysis Transforms 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 Dimensionality Reduction by Inﬁnite-Order Statistics-Based Components Analysis Transforms 6.4.1 6.4.2 6.4.3 Dimensionality Reduction by Projection Pursuit-Based Components Analysis Transforms 6.5.1 6.5.2 6.5.3 6.5.4 Dimensionality Reduction by Feature Extraction-Based Transforms 6.6.1 6.6.2 Dimensionality Reduction by Band Selection Projection Index-Based Projection Pursuit Random Projection Index-Based Projection Pursuit Projection Index-Based Prioritized Projection Pursuit Initialization Driven Projection Pursuit Fisher’s Linear Discriminant Analysis Orthogonal Subspace Projection 135 138 140 140 142 143 144 144 145 146 149 155 163 168 168 170 170 170 172 174 176 176 177 179 179 181 182 182 183 184 187 188 189 190 191 192 193 194 195 195 196 196

资料库

Hyperspectral Data Processing_ Algorithm Design and Analysis.pdf

相关推荐

人工智能

热门标签

最新资料