Bayesian estimation of regularization and PSF parameters for Wiener-Hunt deconvolution

  title={Bayesian estimation of regularization and PSF parameters for Wiener-Hunt deconvolution},
  author={François Orieux and Jean-François Giovannelli and Thomas Rodet},
  journal={Journal of the Optical Society of America. A, Optics, image science, and vision},
  volume={27 7},
  • F. Orieux, J. Giovannelli, T. Rodet
  • Published 30 April 2010
  • Computer Science, Mathematics, Physics, Medicine
  • Journal of the Optical Society of America. A, Optics, image science, and vision
This paper tackles the problem of image deconvolution with joint estimation of point spread function (PSF) parameters and hyperparameters. Within a Bayesian framework, the solution is inferred via a global a posteriori law for unknown parameters and object. The estimate is chosen as the posterior mean, numerically calculated by means of a Monte Carlo Markov chain algorithm. The estimates are efficiently computed in the Fourier domain, and the effectiveness of the method is shown on simulated… 
Semi-unsupervised Bayesian convex image restoration with location mixture of Gaussian
An incremental contribution about a Bayesian approach where a convex field is constructed via Location Mixture of Gaussian and the estimator computed with a fast MCMC algorithm and results have equivalent quality in a quasi-unsupervised approach and go with uncertainty quantification.
Unsupervised joint deconvolution and segmentation method for textured images: a Bayesian approach and an advanced sampling algorithm
The method relies on a hierarchical model and a Bayesian strategy to jointly estimate the labels, the K textured images in addition to hyperparameters: the signal and the noise levels as well as the texture parameters and the granularity coefficient.
Estimating hyperparameters and instrument parameters in regularized inversion. Illustration for SPIRE/Herschel map making
A Bayesian framework that is based on the \post density for all unknown quantities, given the observations is developed, which is explored by a Markov Chain Monte-Carlo sampling technique based on a Gibbs loop and including a Metropolis-Hastings step.
Approximate Bayesian computation for machine learning, inverse problems and big data
This paper summarizes my tutorial talk in MaxEnt 2016 workshop. Starting from the basics of the Bayesian approach and simple example of low dimensional parameter estimation where almost all the
PSF homogenization for multi-band photometry from space on extended objects
We present a fast and robust tool to create PSF matching kernels for multi-band photometric studies. Such kernels are useful for convolving images down to the same PSF prior to computing flux
Fast Joint Multiband Reconstruction From Wideband Images Based on Low-Rank Approximation
This work proposes a new approach combining a precise instrument model for the degraded multispectral images together with a spectral approximation on a low-rank subspace of the object to address the ill-posed inverse problem of reconstruction from wideband images.
Noise bias compensation for tone mapped noisy image using prior knowledge
This work proposes a new denoising method for a tone mapped noisy image, and introduces the prior knowledge, not on the local patch in NLM filters but, on the noise bias (NB) which has not been utilized so far.
Bayesian Estimation for Optimized Structured Illumination Microscopy
This work presents a new framework based on a Bayesian inverse problem formulation approach that enables the computation of one HR image from a reduced number of LR images and has no specific constraints on the modulation to represent a decisive advance for a wider use of HR microscopy through structured illumination.
Fast myopic 2D-SIM Super Resolution Microscopy with Joint Modulation Pattern Estimation
Super-resolution in Structured Illumination Microscopy (SIM) is obtained through de-aliasing of modulated raw images, in which high frequencies are measured indirectly inside the optical transfer
Simultaneous recovery of both bright and dim structures from noisy fluorescence microscopy images using a modified TV constraint
This work modify the general constrained reconstruction method so that both bright and dim structures can drive the deconvolution with equal force, and can simultaneously obtain high‐quality reconstruction across a wide range of signals within a single image or image sequence.


Blind Deconvolution Using a Variational Approach to Parameter, Image, and Blur Estimation
It is shown how the gamma distributions on the unknown hyperparameters can be used to prevent the proposed blind deconvolution method from converging to undesirable image and blur estimates and also how these distributions can be inferred in realistic situations.
Hyperparameter estimation for satellite image restoration using a MCMC maximum-likelihood method
A Markov chain Monte Carlo maximum-likelihood (MCMCML) technique is presented, able to simultaneously achieve the estimation and the reconstruction of satellite image deconvolution.
Unsupervised Bayesian Convex Deconvolution Based on a Field With an Explicit Partition Function
  • J. Giovannelli
  • Mathematics, Medicine
    IEEE Transactions on Image Processing
  • 2008
This paper proposes a non-Gaussian Markov field with a special feature: an explicit partition function that enables the development of an unsupervised edge-preserving convex deconvolution method.
A variational approach for Bayesian blind image deconvolution
This paper presents a new methodology based on a variational approximation, which has been recently introduced for several machine learning problems, and can be viewed as a generalization of the expectation maximization (EM) algorithm.
Iterative statistical approach to blind image deconvolution
  • Lam, Goodman
  • Computer Science, Physics
    Journal of the Optical Society of America. A, Optics, image science, and vision
  • 2000
This work employs a statistical point of view of the data and uses a modified maximum a posteriori approach to identify the most probable object and blur given the observed image to enhance the deconvolution results.
Hierarchical Bayesian Sparse Image Reconstruction With Application to MRFM
A hierarchical Bayesian model to reconstruct sparse images when the observations are obtained from linear transformations and corrupted by an additive white Gaussian noise is presented and a Gibbs sampling strategy is proposed to overcome the complexity of the posterior distribution.
A generalized Gaussian image model for edge-preserving MAP estimation
The authors present a Markov random field model which allows realistic edge modeling while providing stable maximum a posterior (MAP) solutions and is demonstrated to be useful for image reconstruction in low-dosage transmission tomography.
Estimation of Markov random field prior parameters using Markov chain Monte Carlo maximum likelihood
This work details the theory required, and presents an algorithm that is easily implemented and practical in terms of computation time, and demonstrates this algorithm on three MRF models--the standard Potts model, an inhomogeneous variation of the Pottsmodel, and a long-range interaction model better adapted to modeling real-world images.
Strict a priori constraints for maximum-likelihood blind deconvolution
We present a maximum-likelihood approach to improve blind deconvolution of an image. Blind deconvolution is performed through the minimization of an error function by use of the conjugate gradient
Image reconstruction and restoration: overview of common estimation structures and problems
  • G. Demoment
  • Mathematics, Computer Science
    IEEE Trans. Acoust. Speech Signal Process.
  • 1989
The problem of image reconstruction and restoration is first formulated, and some of the current regularization approaches used to solve the problem are described, and a Bayesian interpretation of the regularization techniques is given.