Multi-frame image super-resolution (SR) aims to combine the sub-pixel information from a sequence of low-resolution (LR) images to build a highresolution (HR) one. SR techniques usually suffers from annoying restoration artifacts such as noise, jagged edges, and staircasing effect. In this paper, we aim to increase the performance of SR reconstitution under a variational framework using adaptive diffusion-based regularization term. We propose a new tensor based diffusion regularization that takes the benefit from the diffusion model of Perona-Malik in the flat regions and use a nonlinear tensor derived from the diffusion process of weickert filter near boundaries. Thus, the proposed SR approach can preserve important image features (sharp edges and corners) much better while avoiding artifacts. The synthetic and real experimental results show the effectiveness of the proposed regularisation compared to other methods in both quantitatively and visually.
We consider the Blind Source Separation (BSS) problem in the noisy context. We propose a new methodology in order to enhance separation performances in terms of efficiency and robustness. Our approach consists in denoising the observed signals through the minimization of their total variation, and then minimizing divergence separation criteria combined with the total variation of the estimated source signals. We show by the way that the method leads to some projection problems that are solved by means of projected gradient algorithms. The effeciency and robustness of the proposed algorithm using Hellinger divergence, are illustrated and compared with the classical mutual information approach, through numerical simulations.
International audienceThe aim of a Super Resolution (SR) technique is to construct a high-resolution image from a sequence of observed low-resolution ones of the same scene. One major roadblock of an SR reconstitution is removing noise and blur without destroying edges. We propose a novel multiframe image SR algorithm based on a convex combination of Bilateral Total Variation and a non-smooth second order variational regulariza-tion, using a controlled weighting parameter. We prove the existence of a minimizer of the proposed SR model in the space of functions of bounded Hessian, and we confirm the success of this approach in avoiding undesirable artifacts. The simulation results show the efficient performance of the proposed algorithm compared to other methods in the literature using two criteria, PSNR and SSIM
In this work, we propose a new method to enhance text in document-image. Firstly, we introduce a classical model and a way to solve it by means of a non-convex optimization problem. So, a simoultaneaous estimation of the reflectance and the luminance is obtained when the non uniform illumination (also called luminance) is a smooth function and the reflectance is a function of bounded variation. We give an analyse of this problem and some conditions of existence and unicity. Then, we consider the "log" of the classical model. A new pde's model is proposed. This method is based on the resolution of an original partial differential equation (PDE) estimating the log of the luminance. We assume that the luminance is enough smooth and is the solution of a non classical second order's PDE.Then we deduce the reflectance from the estimated luminance and the acquired image. The effectiveness and the robustness of the proposed process are shown on numerical examples in real-world situation (images acquired from cameraphones). Then, we illustrate the ability of this method to improve an Optical Character Recognition (OCR) in text recognition.
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.