Blind image deconvolution : theory and applications /

Blind image deconvolution : theory and applications /

Saved in:
Bibliographic Details
Imprint:Boca Raton : CRC Press, c2007.
Description:448 p. : ill. (some col.) ; 25 cm.
Language:English
Subject:
Image processing -- Digital techniques -- Mathematics.
Spectrum analysis -- Deconvolution -- Mathematics.
Format: Print Book
URL for this record:http://pi.lib.uchicago.edu/1001/cat/bib/7247389
Hidden Bibliographic Details
Other authors / contributors:Campisi, Patrizio, 1968-
Egiazarian, K. (Karen), 1959-
ISBN:9780849373671 (alk. paper)
0849373670 (alk. paper)
Notes:Includes bibliographical references and index.
committed to retain 20170930 20421213 HathiTrust
Table of Contents:
  • 1. Blind Image Deconvolution: Problem Formulation and Existing Approaches
  • 1.1. Introduction
  • 1.2. Mathematical Problem Formulation
  • 1.3. Classification of Blind Image Deconvolution Methodologies
  • 1.4. Bayesian Framework for Blind Image Deconvolution
  • 1.5. Bayesian Modeling of Blind Image Deconvolution
  • 1.5.1. Observation Model
  • 1.5.2. Parametric Prior Blur Models
  • 1.5.3. Prior Image and Blur Models
  • 1.5.4. Hyperprior Models
  • 1.6. Bayesian Inference Methods in Blind Image Deconvolution
  • 1.6.1. Maximum a Posteriori and Maximum Likelihood
  • 1.6.2. Minimum Mean Squared Error
  • 1.6.3. Marginalizing Hidden Variables
  • 1.6.4. Variational Bayesian Approach
  • 1.6.5. Sampling Methods
  • 1.7. Non-Bayesian Blind Image Deconvolution Models
  • 1.7.1. Spectral and Cepstral Zero Methods
  • 1.7.2. Zero Sheet Separation Algorithms
  • 1.7.3. ARMA Parameter Estimation Algorithms
  • 1.7.4. Nonparametric Deterministic Constraints Algorithms
  • 1.7.5. Nonparametric Algorithms Based on Higher-Order Statistics
  • 1.7.6. Total Least Squares (TLS)
  • 1.7.7. Learning-Based Algorithms
  • 1.7.8. Methods for Spatially Varying Degradation
  • 1.7.9. Multichannel Methods
  • 1.8. Conclusions
  • References
  • 2. Blind Image Deconvolution Using Bussgang Techniques: Applications to Image Deblurring and Texture Synthesis
  • Abstract
  • 2.1. Introduction
  • 2.2. Bussgang Processes
  • 2.3. Single-Channel Bussgang Deconvolution
  • 2.3.1. Convergency Issue
  • 2.3.2. Application to Texture Synthesis
  • 2.4. Multichannel Bussgang Deconvolution
  • 2.4.1. The Observation Model
  • 2.4.2. Multichannel Wiener Filter
  • 2.4.3. Multichannel Bussgang Algorithm
  • 2.4.4. Application to Image Deblurring: Binary Text and Spiky Images
  • 2.4.5. Application to Image Deblurring: Natural Images
  • 2.5. Conclusions
  • References
  • 3. Blind Multiframe Image Deconvolution Using Anisotropic Spatially Adaptive Filtering for Denoising and Regularization
  • Abstract
  • 3.1. Introduction
  • 3.1.1. Blind and Nonblind Inverse
  • 3.1.2. Inverse Regularization
  • 3.2. Observation Model and Preliminaries
  • 3.3. Frequency Domain Equations
  • 3.4. Projection Gradient Optimization
  • 3.5. Anisotropic LPA-ICI Spatially Adaptive Filtering
  • 3.5.1. Motivation
  • 3.5.2. Sectorial Neighborhoods
  • 3.5.3. Adaptive Window Size
  • 3.5.4. LPA-ICI Filtering
  • 3.6. Blind Deconvolution Algorithm
  • 3.6.1. Main Procedure
  • 3.6.2. Image Alignment
  • 3.7. Identifiability and Convergence
  • 3.7.1. Perfect Reconstruction
  • 3.7.2. Hessian and Identifiability
  • 3.7.3. Conditioning and Convergence Rate
  • 3.8. Simulations
  • 3.8.1. Criteria and Algorithm Parameters
  • 3.8.2. Illustrative Results
  • 3.8.3. Perfect Reconstruction
  • 3.8.4. Numerical Results
  • 3.8.5. Image Alignment
  • 3.8.6. Reconstruction of Color Images
  • 3.9. Conclusions
  • Acknowledgments
  • References
  • 4. Bayesian Methods Based on Variational Approximations for Blind Image Deconvolution
  • Abstract
  • 4.1. Introduction
  • 4.2. Background on Variational Methods
  • 4.3. Variational Blind Deconvolution
  • 4.3.1. Variational Functional F(q, ?)
  • 4.3.2. Maximization of the Variational Bound F(q, ?)
  • 4.4. Numerical Experiments
  • 4.4.1. Partially Known Case
  • 4.4.2. Unknown Case
  • 4.5. Conclusions and Future Work
  • Appendix A. Computation of the Variational Bound F(q, ?)
  • Appendix B. Maximization of F{{q, ?)
  • References
  • 5. Deconvolution of Medical Images from Microscopic to Whole Body Images
  • Abstract
  • 5.1. Introduction
  • 5.1.1. Medical Imaging: Tendencies and Goals
  • 5.1.2. Linear Modeling of Image Formation
  • 5.1.3. Blind Deconvolution in Medical Ultrasound Imaging
  • 5.1.4. Blind Deconvolution in Single Photon Emission Computed Tomography
  • 5.1.5. Blind Deconvolution in Confocal Microscopy
  • 5.1.6. Organization of the Chapter
  • 5.2. Nonblind Deconvolution
  • 5.2.1. Regularization via Maximum a Posteriori Estimation
  • 5.2.2. Numerical Optimization via Newton Method
  • 5.2.3. Blind Deconvolution with Shift-Variant Blurs
  • 5.3. Blind Deconvolution in Ultrasound Imaging
  • 5.3.1. Blind Deconvolution via Statistical Modeling
  • 5.3.2. Blind Deconvolution via Higher-Order Spectra Analysis
  • 5.3.3. Horaomorphic Deconvolution: 1-D Case
  • 5.3.4. Homomorphic Deconvolution: 2-D Case
  • 5.3.5. Generalized Homomorphic Deconvolution
  • 5.3.6. Blind Deconvolution via Inverse Filtering
  • 5.4. Blind Deconvolution in Spect
  • 5.4.1. Origins of the Blurring Artifact in Spect
  • 5.4.2. Blind Deconvolution via Alternative Minimization
  • 5.4.3. Blind Deconvolution via Nonnegativity and Support Constrains Recursive Inverse Filtering
  • 5.5. Blind Deconvolution in Confocal Microscopy
  • 5.5.1. Maximum Likelihood Deconvolution in Fluorescence Microscopy
  • 5.5.2. Refinements of the EM Algorithms
  • 5.5.3. Blind Deconvolution in 3-D Transmitted Light Brightfield Microscopy
  • 5.6. Summary
  • References
  • 6. Bayesian Estimation of Blur and Noise in Remote Sensing Imaging
  • Abstract
  • 6.1. Introduction
  • 6.1.1. Blind Deconvolution: State of the Art
  • 6.1.2. Constraining a Difficult Problem
  • 6.1.3. The Bayesian Viewpoint
  • 6.2. The Forward Model
  • 6.2.1. Modeling the Natural Scene Using Fractals
  • 6.2.2. Understanding the Image Formation
  • 6.3. Bayesian Estimation: Invert the Forward Model
  • 6.3.1. Marginalization and Related Approximations
  • 6.3.2. A Natural Parameter Estimation Algorithm (BLINDE)
  • 6.3.3. Why Use a Simplified Model?
  • 6.3.4. A Simplified, Optimized Algorithm
  • 6.4. Possible Improvements and Further Development
  • 6.4.1. Computing Uncertainties
  • 6.4.2. Model Assessment and Checking
  • 6.4.3. Robustness-Related Improvements
  • 6.5. Results
  • 6.5.1. First Method: Blinde
  • 6.5.2. Second method
  • 6.6. Conclusions
  • Acknowledgments
  • References
  • 7. Deconvolution and Blind Deconvolution in Astronomy
  • Abstract
  • 7.1. Introduction
  • 7.2. The Deconvolution Problem
  • 7.3. Linear Regularized Methods
  • 7.3.1. Least Squares Solution
  • 7.3.2. Tikhonov Regularization
  • 7.4. Clean
  • 7.5. Bayesian Methodology
  • 7.5.1. Definition
  • 7.5.2. Maximum Likelihood with Gaussian Noise
  • 7.5.3. Gaussian Bayes Model
  • 7.5.4. Maximum Likelihood with Poisson Noise
  • 7.5.5. Maximum a Posteriori with Poisson Noise
  • 7.5.6. Maximum Entropy Method
  • 7.5.7. Other Regularization Models
  • 7.6. Iterative Regularized Methods
  • 7.6.1. Constraints
  • 7.6.2. Jansson-Van Cittert Method
  • 7.6.3. Other Iterative Methods
  • 7.7. Wavelet-Based Deconvolution
  • 7.7.1. Introduction
  • 7.7.2. Regularization from the Multiresolution Support
  • 7.7.3. Multiresolution Clean
  • 7.7.4. The Wavelet Constraint
  • 7.8. Deconvolution and Resolution
  • 7.9. Myopic and Blind Deconvolution
  • 7.9.1. Myopic Deconvolution
  • 7.9.2. Blind Deconvolution
  • 7.10. Conclusions and Chapter Summary
  • Acknowledgments
  • References
  • 8. Multiframe Blind Deconvolution Coupled with FrameRegistration and Resolution Enhancement
  • Abstract.
  • 8.1. Introduction
  • 8.2. Mathematical Model
  • 8.3. Polyphase Formulation
  • 8.3.1. Integer Downsampling Factor
  • 8.3.2. Rational Downsampling Factor
  • 8.4. Reconstruction of Volatile Blurs
  • 8.4.1. The MBD Case
  • 8.4.2. The RSR Case
  • 8.5. Blind Superresolution
  • 8.6. Experiments
  • 8.6.1. Simulated Data
  • 8.6.2. Real Data
  • 8.6.3. Performance Experiments
  • 8.7. Conclusions
  • Acknowledgment
  • References
  • 9. Blind Reconstruction of Multiframe Imagery Based on Fusion and Classification
  • Abstract.
  • 9.1. Introduction
  • 9.2. System Overview
  • 9.3. Recursive Inverse Filtering with Finite Normal-Density Mixtures (RIF-FNM)
  • 9.3.1. Image Modeling Using Finite Mixture Distributions
  • 9.3.2. Pixel Classification
  • 9.3.3. ML-Based Image Fusion
  • 9.4. Optimal Filter Adaptation
  • 9.5. Effects of Noise
  • 9.6. The Fusion and Classification Recursive Inverse Filtering Algorithm (FAC-RIF)
  • 9.6.1. The Iterative Algorithm
  • 9.6.2. Prefiltering and Postfiltering Processing
  • 9.6.3. Classification
  • 9.6.4. Fusion-Based Classification
  • 9.6.5. Fusion of Reconstructed Images
  • 9.7. Experimental Results
  • 9.8. Final Remarks
  • References
  • 10. Blind Deconvolution and Structured Matrix Computations with Applications to Array Imaging
  • Abstract
  • 10.1. Introduction
  • 10.2. One-Dimensional Deconvolution Formulation
  • 10.3. Regularized and Constrained TLS Formulation
  • 10.3.1. Symmetric Point Spread Functions
  • 10.4. Numerical Algorithms
  • 10.4.1. The Preconditioned Conjugate Gradient Method
  • 10.4.2. Cosine Transform-Based Preconditioners
  • 10.5. Two-Dimensional Deconvolution Problems
  • 10.6. Numerical Examples
  • 10.7. Application: High-Resolution Image Reconstruction
  • 10.7.1. Mathematical Model
  • 10.7.2. Image Reconstruction Formulation
  • 10.7.3. Simulation Results
  • 10.8. Concluding Remarks and Current Work
  • Acknowledgments
  • References
  • Index
Gift of the Clinton M. and Dorothy R. Doede Book Fund bookplate
The John Crerar Library bookplate

Similar Items