Approximation and Optimization in Image Restoration and Reconstruction
نویسندگان
چکیده
Recent studies in image denoising show the importance of using (i) multiresolution transformations with improved orientation selectivity and (ii) the appropriate spatial context models. Regarding the first point, a number of the so-called " geometrical representations " have been introduced that represent image discontinuities better than the classical wavelets, which also results in better noise reduction results. It has been also proved that using better models for the statistical dependencies of the transform coefficients improves the estimation of the noise free data. A number of different contextual models have been proposed in this respect starting from the simple bivariate parent-child models or local spatial indicators to the more complex Markov Random Field (MRF) models, Gaussian Scale Mixture (GSM) models and recently mixtures of the GSM (MGSM), the mixtures of the projected GSM (MPGSM), Field of Experts (FoE) models and Gaussian Conditional Random Field (GCRF) models. In this talk, we wish to discuss these developments from the perspective of our own research and to identify possible directions for future developments in this field. In particular, we will discuss detection and estimation of a signal of interest in the noisy image using a local spatial activity indicator, a MRF model and a GSM model. We will then compare this approach with a state-of-the art estimator using MPGSM model and link it to some other most recent developments in this field. We will also show some practical applications to different types of images with natural noise, including some medical imaging modalities and digital camera images.
منابع مشابه
Presentation summaries Approximation and Optimization in Image Restoration and Reconstruction
متن کامل
Image Restoration by Variable Splitting based on Total Variant Regularizer
The aim of image restoration is to obtain a higher quality desired image from a degraded image. In this strategy, an image inpainting method fills the degraded or lost area of the image by appropriate information. This is performed in such a way so that the obtained image is undistinguishable for a casual person who is unfamiliar with the original image. In this paper, different images are degr...
متن کاملRestoration of blurred star field images by maximally sparse optimization
The problem of removing blur from, or sharpening, astronomical star field intensity images is discussed. An approach to image restoration that recovers image detail using a constrained optimization theoretic approach is introduced. Ideal star images may be modeled as a few point sources in a uniform background. It is argued that a direct measure of image sparseness is the appropriate optimizati...
متن کاملApproximation and Optimization in Image Restoration and Reconstruction
Iteratively Re-weighted Least Squares minimization (IRLS) appears for the first time in the approximation practice in the Ph.D. thesis of C. L. Lawson in 1961 for L∞ minimization. In the 1970s extensions of Lawson’s algorithm for `p-minimization were proposed, as reported in the work of M. R. Osborne. IRLS has been proposed for sparse recovery in signal processing in [5] and for total variation...
متن کاملApproximation and Optimization in Image Restoration and Reconstruction
We will present a convex representation for the minimal partition problem, based on the notion of “paired calibrations” [2], [3]. We propose an efficient algorithm for minimizing this problem, and discuss the practical implementation with “many” labels. It can be applied to a number of problems, from segmentation to stereo reconstruction. The quality of the results is substantially better than ...
متن کاملApproximation and Optimization in Image Restoration and Reconstruction
Positron emission tomography (PET) is a biomedical imaging technique that allows for the visualization and the measurement of biological processes. PET is based on the injection to the patient of a molecular probe for the process of interest. The probe is labeled with radio-isotopes that decay by emitting a positron. The emitted positron annihilates almost immediately with an electron, resultin...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2009