Abstract:We propose a new two-phase method for reconstruction of blurred images corrupted by impulse noise. In the first phase, we use a noise detector to identify the pixels that are contaminated by noise, and then, in the second phase, we reconstruct the noisy pixels by solving an equality constrained total variation minimization problem that preserves the exact values of the noise-free pixels. For images that are only corrupted by impulse noise (i.e., not blurred) we apply the semismooth Newton's method to a reduced… Show more
“…Similarly as the two-phase methods described in [3,6,7,30], it may be attractive to preprocess images that are contaminated by impulse noise by a median filter. This would result in a fully automatic two-phase method.…”
Section: Conclusion and Extensionmentioning
confidence: 99%
“…Several other methods have been developed for the restoration of images that are corrupted by blur and impulse noise. In particular, two-phase strategies have been shown to yield accurate restorations; see, e.g., [3,6,7,30]. These methods first identify pixels that are contaminated by impulse noise by means of a median-type filter.…”
Section: Introductionmentioning
confidence: 99%
“…In the current literature on two-phase methods, the functional to be minimized in the second phase is usually convex and little attention is given to the selection of the regularization parameter. In fact, only Sciacchitano et al [30] propose a two-phase method that does not require a user to specify a regularization parameter. None of the methods mentioned is designed to remove mixed noise, i.e., noise that is made up of both impulse noise and Gaussian noise.…”
Discrete ill-posed problems arise in many areas of science and engineering. Their solutions, if they exist, are very sensitive to perturbations in the data. Regularization aims to reduce this sensitivity. Many regularization methods replace the original problem by a minimization problem with a fidelity term and a regularization term. Recently, the use of a p-norm to measure the fidelity term and a q-norm to measure the regularization term has received considerable attention. The relative importance of these terms is determined by a regularization parameter. When the perturbation in the available data is made up of impulse noise and a sparse solution is desired, it often is beneficial to let 0 < p, q < 1. Then the p-and q-norms are not norms. The choice of a suitable regularization parameter is crucial for the quality of the computed solution. It therefore is important to develop methods for determining this parameter automatically, without user-interaction. However, the latter has so far not received much attention when the data is contaminated by impulse noise. This paper discusses two approaches based on cross validation for determining the regularization parameter in this situation. Computed examples that illustrate the performance of these approaches when applied to the restoration of impulse noise contaminated images are presented.
“…Similarly as the two-phase methods described in [3,6,7,30], it may be attractive to preprocess images that are contaminated by impulse noise by a median filter. This would result in a fully automatic two-phase method.…”
Section: Conclusion and Extensionmentioning
confidence: 99%
“…Several other methods have been developed for the restoration of images that are corrupted by blur and impulse noise. In particular, two-phase strategies have been shown to yield accurate restorations; see, e.g., [3,6,7,30]. These methods first identify pixels that are contaminated by impulse noise by means of a median-type filter.…”
Section: Introductionmentioning
confidence: 99%
“…In the current literature on two-phase methods, the functional to be minimized in the second phase is usually convex and little attention is given to the selection of the regularization parameter. In fact, only Sciacchitano et al [30] propose a two-phase method that does not require a user to specify a regularization parameter. None of the methods mentioned is designed to remove mixed noise, i.e., noise that is made up of both impulse noise and Gaussian noise.…”
Discrete ill-posed problems arise in many areas of science and engineering. Their solutions, if they exist, are very sensitive to perturbations in the data. Regularization aims to reduce this sensitivity. Many regularization methods replace the original problem by a minimization problem with a fidelity term and a regularization term. Recently, the use of a p-norm to measure the fidelity term and a q-norm to measure the regularization term has received considerable attention. The relative importance of these terms is determined by a regularization parameter. When the perturbation in the available data is made up of impulse noise and a sparse solution is desired, it often is beneficial to let 0 < p, q < 1. Then the p-and q-norms are not norms. The choice of a suitable regularization parameter is crucial for the quality of the computed solution. It therefore is important to develop methods for determining this parameter automatically, without user-interaction. However, the latter has so far not received much attention when the data is contaminated by impulse noise. This paper discusses two approaches based on cross validation for determining the regularization parameter in this situation. Computed examples that illustrate the performance of these approaches when applied to the restoration of impulse noise contaminated images are presented.
“…Refs. [1,2,4,7,9,10,13,17,21,24,26,27,34,35,40,42,43,46]. Some of them, such as wavelet method [37,47] and the compressive sensing method [39,48,49], are also used in traffic analysis.…”
An easily implementable noise strength estimation algorithm to analyse the urban traffic is developed. It improves the accuracy of the road velocity predictions. Using real urban traffic data from Beijing Taxi GPS system, we demonstrate the efficiency of the algorithm. It is also shown that the BV denoising method with the best noisestrength estimates significantly improves the road clustering.
“…[5,23]. On the other hand the total variation (TV) denoising scheme proposed by Rudin et al [19], initiated the development of numerous PDE variational filtering techniques [1,3,7,8,11,15,20,22].…”
A second-order nonlinear anisotropic diffusion-based model for Gaussian additive noise removal is proposed. The method is based on a properly constructed edgestopping function and provides an efficient detail-preserving denoising. It removes additive noise, overcomes blurring effect, reduces the image staircasing and does not generate multiplicative noise, thus preserving boundaries and all the essential image features very well. The corresponding PDE model is solved by a robust finite-difference based iterative scheme consistent with the diffusion model. The method converges very fast to the model solution, the existence and regularity of which is rigorously proved.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.