Available:*
Library | Item Barcode | Call Number | Material Type | Item Category 1 | Status |
---|---|---|---|---|---|
Searching... | 30000010133161 | TA1632 B54 2007 | Open Access Book | Book | Searching... |
On Order
Summary
Summary
Blind image deconvolution is constantly receiving increasing attention from the academic as well the industrial world due to both its theoretical and practical implications. The field of blind image deconvolution has several applications in different areas such as image restoration, microscopy, medical imaging, biological imaging, remote sensing, astronomy, nondestructive testing, geophysical prospecting, and many others. Blind Image Deconvolution: Theory and Applications surveys the current state of research and practice as presented by the most recognized experts in the field, thus filling a gap in the available literature on blind image deconvolution.
Explore the gamut of blind image deconvolution approaches and algorithms that currently exist and follow the current research trends into the future. This comprehensive treatise discusses Bayesian techniques, single- and multi-channel methods, adaptive and multi-frame techniques, and a host of applications to multimedia processing, astronomy, remote sensing imagery, and medical and biological imaging at the whole-body, small-part, and cellular levels. Everything you need to step into this dynamic field is at your fingertips in this unique, self-contained masterwork.
For image enhancement and restoration without a priori information, turn to Blind Image Deconvolution: Theory and Applications for the knowledge and techniques you need to tackle real-world problems.
Table of Contents
1 Blind Image Deconvolution: Problem Formulation and Existing Approaches | p. 1 |
1.1 Introduction | p. 1 |
1.2 Mathematical Problem Formulation | p. 4 |
1.3 Classification of Blind Image Deconvolution Methodologies | p. 5 |
1.4 Bayesian Framework for Blind Image Deconvolution | p. 6 |
1.5 Bayesian Modeling of Blind Image Deconvolution | p. 7 |
1.5.1 Observation Model | p. 7 |
1.5.2 Parametric Prior Blur Models | p. 8 |
1.5.3 Prior Image and Blur Models | p. 9 |
1.5.4 Hyperprior Models | p. 15 |
1.6 Bayesian Inference Methods in Blind Image Deconvolution | p. 16 |
1.6.1 Maximum a Posteriori and Maximum Likelihood | p. 16 |
1.6.2 Minimum Mean Squared Error | p. 18 |
1.6.3 Marginalizing Hidden Variables | p. 19 |
1.6.4 Variational Bayesian Approach | p. 21 |
1.6.5 Sampling Methods | p. 22 |
1.7 Non-Bayesian Blind Image Deconvolution Models | p. 24 |
1.7.1 Spectral and Cepstral Zero Methods | p. 25 |
1.7.2 Zero Sheet Separation Algorithms | p. 26 |
1.7.3 ARMA Parameter Estimation Algorithms | p. 26 |
1.7.4 Nonparametric Deterministic Constraints Algorithms | p. 27 |
1.7.5 Nonparametric Algorithms Based on Higher-Order Statistics | p. 29 |
1.7.6 Total Least Squares (TLS) | p. 29 |
1.7.7 Learning-Based Algorithms | p. 30 |
1.7.8 Methods for Spatially Varying Degradation | p. 30 |
1.7.9 Multichannel Methods | p. 31 |
1.8 Conclusions | p. 32 |
References | p. 32 |
2 Blind Image Deconvolution Using Bussgang Techniques: Applications to Image Deblurring and Texture Synthesis | p. 43 |
Abstract | p. 43 |
2.1 Introduction | p. 44 |
2.2 Bussgang Processes | p. 46 |
2.3 Single-Channel Bussgang Deconvolution | p. 48 |
2.3.1 Convergency Issue | p. 51 |
2.3.2 Application to Texture Synthesis | p. 55 |
2.4 Multichannel Bussgang Deconvolution | p. 61 |
2.4.1 The Observation Model | p. 62 |
2.4.2 Multichannel Wiener Filter | p. 63 |
2.4.3 Multichannel Bussgang Algorithm | p. 64 |
2.4.4 Application to Image Deblurring: Binary Text and Spiky Images | p. 67 |
2.4.5 Application to Image Deblurring: Natural Images | p. 73 |
2.5 Conclusions | p. 85 |
References | p. 88 |
3 Blind Multiframe Image Deconvolution Using Anisotropic Spatially Adaptive Filtering for Denoising and Regularization | p. 95 |
Abstract | p. 95 |
3.1 Introduction | p. 96 |
3.1.1 Blind and Nonblind Inverse | p. 96 |
3.1.2 Inverse Regularization | p. 98 |
3.2 Observation Model and Preliminaries | p. 100 |
3.3 Frequency Domain Equations | p. 102 |
3.4 Projection Gradient Optimization | p. 104 |
3.5 Anisotropic LPA-ICI Spatially Adaptive Filtering | p. 108 |
3.5.1 Motivation | p. 109 |
3.5.2 Sectorial Neighborhoods | p. 110 |
3.5.3 Adaptive Window Size | p. 110 |
3.5.4 LPA-ICI Filtering | p. 111 |
3.6 Blind Deconvolution Algorithm | p. 112 |
3.6.1 Main Procedure | p. 112 |
3.6.2 Image Alignment | p. 114 |
3.7 Identifiability and Convergence | p. 114 |
3.7.1 Perfect Reconstruction | p. 114 |
3.7.2 Hessian and Identifiability | p. 115 |
3.7.3 Conditioning and Convergence Rate | p. 117 |
3.8 Simulations | p. 120 |
3.8.1 Criteria and Algorithm Parameters | p. 120 |
3.8.2 Illustrative Results | p. 122 |
3.8.3 Perfect Reconstruction | p. 128 |
3.8.4 Numerical Results | p. 128 |
3.8.5 Image Alignment | p. 130 |
3.8.6 Reconstruction of Color Images | p. 132 |
3.9 Conclusions | p. 134 |
Acknowledgments | p. 136 |
References | p. 136 |
4 Bayesian Methods Based on Variational Approximations for Blind Image Deconvolution | p. 141 |
Abstract | p. 141 |
4.1 Introduction | p. 142 |
4.2 Background on Variational Methods | p. 145 |
4.3 Variational Blind Deconvolution | p. 146 |
4.3.1 Variational Functional F(q, ?) | p. 146 |
4.3.2 Maximization of the Variational Bound F(q, ?) | p. 149 |
4.4 Numerical Experiments | p. 152 |
4.4.1 Partially Known Case | p. 153 |
4.4.2 Unknown Case | p. 153 |
4.5 Conclusions and Future Work | p. 154 |
Appendix A Computation of the Variational Bound F(q, ?) | p. 155 |
Appendix B Maximization of F{{q, ?) | p. 157 |
References | p. 165 |
5 Deconvolution of Medical Images from Microscopic to Whole Body Images | p. 169 |
Abstract | p. 169 |
5.1 Introduction | p. 170 |
5.1.1 Medical Imaging: Tendencies and Goals | p. 170 |
5.1.2 Linear Modeling of Image Formation | p. 171 |
5.1.3 Blind Deconvolution in Medical Ultrasound Imaging | p. 172 |
5.1.4 Blind Deconvolution in Single Photon Emission Computed Tomography | p. 175 |
5.1.5 Blind Deconvolution in Confocal Microscopy | p. 176 |
5.1.6 Organization of the Chapter | p. 178 |
5.2 Nonblind Deconvolution | p. 179 |
5.2.1 Regularization via Maximum a Posteriori Estimation | p. 179 |
5.2.2 Numerical Optimization via Newton Method | p. 182 |
5.2.3 Blind Deconvolution with Shift-Variant Blurs | p. 183 |
5.3 Blind Deconvolution in Ultrasound Imaging | p. 184 |
5.3.1 Blind Deconvolution via Statistical Modeling | p. 185 |
5.3.2 Blind Deconvolution via Higher-Order Spectra Analysis | p. 190 |
5.3.3 Horaomorphic Deconvolution: 1-D Case | p. 192 |
5.3.4 Homomorphic Deconvolution: 2-D Case | p. 199 |
5.3.5 Generalized Homomorphic Deconvolution | p. 205 |
5.3.6 Blind Deconvolution via Inverse Filtering | p. 211 |
5.4 Blind Deconvolution in Spect | p. 215 |
5.4.1 Origins of the Blurring Artifact in Spect | p. 215 |
5.4.2 Blind Deconvolution via Alternative Minimization | p. 217 |
5.4.3 Blind Deconvolution via Nonnegativity and Support Constrains Recursive Inverse Filtering | p. 222 |
5.5 Blind Deconvolution in Confocal Microscopy | p. 223 |
5.5.1 Maximum Likelihood Deconvolution in Fluorescence Microscopy | p. 223 |
5.5.2 Refinements of the EM Algorithms | p. 227 |
5.5.3 Blind Deconvolution in 3-D Transmitted Light Brightfield Microscopy | p. 227 |
5.6 Summary | p. 228 |
References | p. 230 |
6 Bayesian Estimation of Blur and Noise in Remote Sensing Imaging | p. 239 |
Abstract | p. 239 |
6.1 Introduction | p. 240 |
6.1.1 Blind Deconvolution: State of the Art | p. 241 |
6.1.2 Constraining a Difficult Problem | p. 243 |
6.1.3 The Bayesian Viewpoint | p. 244 |
6.2 The Forward Model | p. 245 |
6.2.1 Modeling the Natural Scene Using Fractals | p. 245 |
6.2.2 Understanding the Image Formation | p. 246 |
6.3 Bayesian Estimation: Invert the Forward Model | p. 249 |
6.3.1 Marginalization and Related Approximations | p. 250 |
6.3.2 A Natural Parameter Estimation Algorithm (BLINDE) | p. 251 |
6.3.3 Why Use a Simplified Model? | p. 253 |
6.3.4 A Simplified, Optimized Algorithm | p. 255 |
6.4 Possible Improvements and Further Development | p. 256 |
6.4.1 Computing Uncertainties | p. 256 |
6.4.2 Model Assessment and Checking | p. 257 |
6.4.3 Robustness-Related Improvements | p. 258 |
6.5 Results | p. 259 |
6.5.1 First Method: Blinde | p. 259 |
6.5.2 Second method | p. 259 |
6.6 Conclusions | p. 271 |
Acknowledgments | p. 271 |
References | p. 271 |
7 Deconvolution and Blind Deconvolution in Astronomy | p. 277 |
Abstract | p. 277 |
7.1 Introduction | p. 278 |
7.2 The Deconvolution Problem | p. 280 |
7.3 Linear Regularized Methods | p. 282 |
7.3.1 Least Squares Solution | p. 282 |
7.3.2 Tikhonov Regularization | p. 282 |
7.4 Clean | p. 283 |
7.5 Bayesian Methodology | p. 283 |
7.5.1 Definition | p. 284 |
7.5.2 Maximum Likelihood with Gaussian Noise | p. 284 |
7.5.3 Gaussian Bayes Model | p. 285 |
7.5.4 Maximum Likelihood with Poisson Noise | p. 285 |
7.5.5 Maximum a Posteriori with Poisson Noise | p. 286 |
7.5.6 Maximum Entropy Method | p. 286 |
7.5.7 Other Regularization Models | p. 287 |
7.6 Iterative Regularized Methods | p. 288 |
7.6.1 Constraints | p. 288 |
7.6.2 Jansson-Van Cittert Method | p. 289 |
7.6.3 Other Iterative Methods | p. 289 |
7.7 Wavelet-Based Deconvolution | p. 290 |
7.7.1 Introduction | p. 290 |
7.7.2 Regularization from the Multiresolution Support | p. 292 |
7.7.3 Multiresolution Clean | p. 295 |
7.7.4 The Wavelet Constraint | p. 295 |
7.8 Deconvolution and Resolution | p. 300 |
7.9 Myopic and Blind Deconvolution | p. 301 |
7.9.1 Myopic Deconvolution | p. 303 |
7.9.2 Blind Deconvolution | p. 305 |
7.10 Conclusions and Chapter Summary | p. 308 |
Acknowledgments | p. 309 |
References | p. 309 |
8 Multiframe Blind Deconvolution Coupled with FrameRegistration and Resolution Enhancement | p. 317 |
Abstract p. 317 | |
8.1 Introduction | p. 318 |
8.2 Mathematical Model | p. 321 |
8.3 Polyphase Formulation | p. 323 |
8.3.1 Integer Downsampling Factor | p. 324 |
8.3.2 Rational Downsampling Factor | p. 325 |
8.4 Reconstruction of Volatile Blurs | p. 326 |
8.4.1 The MBD Case | p. 327 |
8.4.2 The RSR Case | p. 327 |
8.5 Blind Superresolution | p. 330 |
8.6 Experiments | p. 333 |
8.6.1 Simulated Data | p. 333 |
8.6.2 Real Data | p. 335 |
8.6.3 Performance Experiments | p. 338 |
8.7 Conclusions | p. 345 |
Acknowledgment | p. 345 |
References | p. 346 |
9 Blind Reconstruction of Multiframe Imagery Based on Fusion and Classification | p. 349 |
Abstract p. 349 | |
9.1 Introduction | p. 350 |
9.2 System Overview | p. 351 |
9.3 Recursive Inverse Filtering with Finite Normal-Density Mixtures (RIF-FNM) | p. 352 |
9.3.1 Image Modeling Using Finite Mixture Distributions | p. 352 |
9.3.2 Pixel Classification | p. 356 |
9.3.3 ML-Based Image Fusion | p. 357 |
9.4 Optimal Filter Adaptation | p. 358 |
9.5 Effects of Noise | p. 360 |
9.6 The Fusion and Classification Recursive Inverse Filtering Algorithm (FAC-RIF) | p. 361 |
9.6.1 The Iterative Algorithm | p. 362 |
9.6.2 Prefiltering and Postfiltering Processing | p. 363 |
9.6.3 Classification | p. 363 |
9.6.4 Fusion-Based Classification | p. 364 |
9.6.5 Fusion of Reconstructed Images | p. 364 |
9.7 Experimental Results | p. 365 |
9.8 Final Remarks | p. 371 |
References | p. 371 |
10 Blind Deconvolution and Structured Matrix Computations with Applications to Array Imaging | p. 377 |
Abstract | p. 377 |
10.1 Introduction | p. 378 |
10.2 One-Dimensional Deconvolution Formulation | p. 379 |
10.3 Regularized and Constrained TLS Formulation | p. 382 |
10.3.1 Symmetric Point Spread Functions | p. 386 |
10.4 Numerical Algorithms | p. 388 |
10.4.1 The Preconditioned Conjugate Gradient Method | p. 390 |
10.4.2 Cosine Transform-Based Preconditioners | p. 394 |
10.5 Two-Dimensional Deconvolution Problems | p. 396 |
10.6 Numerical Examples | p. 398 |
10.7 Application: High-Resolution Image Reconstruction | p. 400 |
10.7.1 Mathematical Model | p. 403 |
10.7.2 Image Reconstruction Formulation | p. 406 |
10.7.3 Simulation Results | p. 411 |
10.8 Concluding Remarks and Current Work | p. 416 |
Acknowledgments | p. 418 |
References | p. 418 |
Index | p. 423 |