Change Detection and Image Time-Series Analysis 1: Unervised Methods

ISBN-13: 9781789450569 / Angielski / Twarda / 2022 / 304 str.

Atto, Abdourrahmane M.

Spis treści

Preface xiAbdourrahmane M. ATTO, Francesca BOVOLO and Lorenzo BRUZZONEList of Notations xvChapter 1. Unsupervised Change Detection in Multitemporal Remote Sensing Images 1Sicong LIU, Francesca BOVOLO, Lorenzo BRUZZONE, Qian DU and Xiaohua TONG1.1. Introduction 11.2. Unsupervised change detection in multispectral images 31.2.1.Related concepts 31.2.2.Open issues and challenges 71.2.3. Spectral-spatial unsupervised CD techniques 71.3. Unsupervised multiclass change detection approaches based on modelling spectral-spatial information 91.3.1. Sequential spectral change vector analysis (S2CVA) 91.3.2. Multiscale morphological compressed change vector analysis 111.3.3. Superpixel-level compressed change vector analysis 151.4.Dataset description and experimental setup 181.4.1.Dataset description 181.4.2.Experimental setup 221.5.Results anddiscussion 241.5.1.Results on the Xuzhou dataset 241.5.2. Results on the Indonesia tsunami dataset 241.6.Conclusion 281.7.Acknowledgements 291.8.References 29Chapter 2. Change Detection in Time Series of Polarimetric SAR Images 35Knut CONRADSEN, Henning SKRIVER, Morton J. CANTY and Allan A. NIELSEN2.1. Introduction 352.1.1.The problem 362.1.2. Important concepts illustrated by means of the gamma distribution 392.2.Test theory and matrix ordering 452.2.1. Test for equality of two complex Wishart distributions 452.2.2. Test for equality of k-complex Wishart distributions 472.2.3. The block diagonal case 492.2.4.The Loewner order 522.3.The basic change detection algorithm 532.4.Applications 552.4.1.Visualizingchanges 582.4.2.Fieldwise change detection 592.4.3. Directional changes using the Loewner ordering 622.4.4. Software availability 652.5.References 70Chapter 3. An Overview of Covariance-based Change Detection Methodologies in Multivariate SAR Image Time Series 73Ammar MIAN, Guillaume GINOLHAC, Jean-Philippe OVARLEZ, Arnaud BRELOY and Frédéric PASCAL3.1. Introduction 733.2.Dataset description 763.3.Statistical modelling of SAR images 773.3.1.The data 773.3.2.Gaussian model 773.3.3.Non-Gaussianmodeling 833.4.Dissimilarity measures 843.4.1.Problem formulation 843.4.2. Hypothesis testing statistics 853.4.3. Information-theoretic measures 873.4.4.Riemannian geometry distances 893.4.5.Optimal transport 903.4.6.Summary 913.4.7. Results of change detectors on the UAVSAR dataset 913.5. Change detection based on structured covariances 943.5.1. Low-rank Gaussian change detector 963.5.2. Low-rank compound Gaussian change detector 973.5.3. Results of low-rank change detectors on the UAVSAR dataset 1003.6.Conclusion 1023.7.References 103Chapter 4. Unsupervised Functional Information Clustering in Extreme Environments from Filter Banks and Relative Entropy 109Abdourrahmane M. ATTO, Fatima KARBOU, Sophie GIFFARD-ROISIN and Lionel BOMBRUN4.1. Introduction 1094.2.Parametric modelling of convnet features 1104.3.Anomaly detection in image time series 1134.4.Functional image time series clustering 1194.5.Conclusion 1234.6.References 123Chapter 5. Thresholds and Distances to Better Detect Wet Snow over Mountains with Sentinel-1 Image Time Series 127Fatima KARBOU, Guillaume JAMES, Philippe DURAND and Abdourrahmane M. ATTO5.1. Introduction 1275.2.Test area and data 1295.3.Wet snowdetectionusingSentinel-1 1295.4.Metrics to detect wet snow 1335.5.Discussion 1385.6.Conclusion 1435.7.Acknowledgements 1435.8.References 143Chapter 6. Fractional Field Image Time Series Modeling and Application to Cyclone Tracking 145Abdourrahmane M. ATTO, Aluísio PINHEIRO, Guillaume GINOLHAC and Pedro MORETTIN6.1. Introduction 1456.2. Random field model of a cyclone texture 1486.2.1.Cyclone texture feature 1496.2.2.Wavelet-based power spectral densities and cyclone fields 1506.2.3. Fractional spectral power decay model 1536.3.Cyclonefield eye detection and tracking 1576.3.1.Cyclone eye detection 1576.3.2.Dynamic fractal field eye tracking 1586.4. Cyclone field intensity evolution prediction 1596.5.Discussion 1616.6.Acknowledgements 1636.7.References 163Chapter 7. Graph of Characteristic Points for Texture Tracking: Application to Change Detection and Glacier Flow Measurement from SAR Images 167Minh-Tan PHAM and Grégoire MERCIER7.1. Introduction 1677.2. Texture representation and characterization using local extrema 1697.2.1.Motivation and approach 1697.2.2. Local extrema keypoints within SAR images 1727.3.Unsupervised change detection 1757.3.1. Proposed framework 1757.3.2.Weighted graph construction from keypoints 1767.3.3.Change measure (CM) generation 1787.4.Experimental study 1797.4.1. Data description and evaluation criteria 1797.4.2.Change detection results 1817.4.3.Sensitivity to parameters 1857.4.4.Comparison with the NLM model 1887.4.5. Analysis of the algorithm complexity 1917.5.Application to glacier flow measurement 1927.5.1. Proposed method 1937.5.2.Results 1947.6.Conclusion 1967.7.References 197Chapter 8. Multitemporal Analysis of Sentinel-1/2 Images for Land Use Monitoring at Regional Scale 201Andrea GARZELLI and Claudia ZOPPETTI8.1. Introduction 2018.2. Proposed method 2038.2.1.Test site anddata 2068.3.SAR processing 2098.4.Optical processing 2158.5.Combination layer 2178.6.Results 2198.7.Conclusion 2208.8.References 221Chapter 9. Statistical Difference Models for Change Detection in Multispectral Images 223Massimo ZANETTI, Francesca BOVOLO and Lorenzo BRUZZONE9.1. Introduction 2239.2. Overview of the change detection problem 2259.2.1. Change detection methods for multispectral images 2279.2.2. Challenges addressed in this chapter 2309.3. The Rayleigh-Rice mixture model for the magnitude of the difference image 2319.3.1. Magnitude image statistical mixture model 2319.3.2.Bayesian decision 2339.3.3. Numerical approach to parameter estimation 2349.4. A compound multiclass statistical model of the difference image 2399.4.1. Difference image statistical mixture model 2409.4.2. Magnitude image statistical mixture model 2459.4.3.Bayesian decision 2489.4.4. Numerical approach to parameter estimation 2499.5.Experimental results 2539.5.1.Dataset description 2539.5.2.Experimental setup 2569.5.3. Test 1: Two-class Rayleigh-Rice mixture model 2569.5.4. Test 2: Multiclass Rician mixture model 2609.6.Conclusion 2669.7.References 267List of Authors 275Index 277Summary of Volume 2 281

O autorze

Abdourrahmane M. Atto is Associate Professor at the University Savoie Mont Blanc, France. His research interests include mathematical methods and models for artificial intelligence and image time series.Francesca Bovolo is the Head of the Remote Sensing for Digital Earth Unit, Fondazione Bruno Kessler, Italy. Her research interests include remote sensing image time series analysis, content-based time series retrieval and radar sounders.Lorenzo Bruzzone is Professor of Telecommunications and the Founder and Director of the Remote Sensing Laboratory at the University of Trento, Italy. His research interests include remote sensing, machine learning and pattern recognition.