Sulam, J., Ophir, B., Elad M, Image Denoising Through Multi-Scale Learnt Dictionaries, International Conference on Image Processing ICIP, November 2014 (oral presentation). (original) (raw)

Image denoising through multi-scale learnt dictionaries

2014 IEEE International Conference on Image Processing (ICIP), 2014

Over the last decade, a number of algorithms have shown promising results in removing additive white Gaussian noise from natural images, and though different, they all share in common a patch based strategy by locally denoising overlapping patches. While this lowers the complexity of the problem, it also causes noticeable artifacts when dealing with large smooth areas. In this paper we present a patch-based denoising algorithm relying on a sparsity-inspired model (K-SVD), which uses a multi-scale analysis framework. This allows us to overcome some of the disadvantages of the popular algorithms. We look for a sparse representation under an already sparsifying wavelet transform by adaptively training a dictionary on the different decomposition bands of the noisy image itself, leading to a multi-scale version of the K-SVD algorithm. We then combine the single scale and multi-scale approaches by merging both outputs by weighted joint sparse coding of the images. Our experiments on natural images indicate that our method is competitive with state of the art algorithms in terms of PSNR while giving superior results with respect to visual quality.

Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries

IEEE Transactions on Image Processing, 2000

We address the image denoising problem, where zero-mean white and homogeneous Gaussian additive noise is to be removed from a given image. The approach taken is based on sparse and redundant representations over trained dictionaries. Using the K-SVD algorithm, we obtain a dictionary that describes the image content effectively. Two training options are considered: using the corrupted image itself, or training on a corpus of high-quality image database. Since the K-SVD is limited in handling small image patches, we extend its deployment to arbitrary image sizes by defining a global image prior that forces sparsity over patches in every location in the image. We show how such Bayesian treatment leads to a simple and effective denoising algorithm. This leads to a state-of-the-art denoising performance, equivalent and sometimes surpassing recently published leading alternative denoising methods.

Multiscale Sparsifying Transform Learning for Image Denoising

ArXiv, 2020

The data-driven sparse methods such as synthesis dictionary learning and sparsifying transform learning have been proven to be effective in image denoising. However, these methods are intrinsically single-scale, which ignores the multiscale nature of images. This often leads to suboptimal results. In this paper, we propose several strategies to exploit multiscale information in image denoising through the sparsifying transform learning denoising (TLD) method. To this end, we first employ a simple method of denoising each wavelet subband independently via TLD. Then, we show that this method can be greatly enhanced using wavelet subbands mixing, which is a cheap fusion technique, to combine the results of single-scale and multiscale methods. Finally, we remove the need for denoising detail subbands. This simplification leads to an efficient multiscale denoising method with competitive performance to its baseline. The effectiveness of the proposed methods are experimentally shown over ...

Image denoising using locally learned dictionaries

Computational Imaging VII, 2009

In this paper we discuss a novel patch-based framework for image denoising through local geometric representations of an image. We learn local data adaptive bases that best capture the underlying geometric information from noisy image patches. To do so we first identify regions of similar structure in the given image and group them together. This is done by the use of meaningful features in the form of local kernels that capture similarities between pixels in a neighborhood. We then learn an informative basis (called a dictionary) for each cluster that best describes the patches in the cluster. Such a data representation can be achieved by performing a simple principal component analysis (PCA) on the member patches of each cluster. The number of principal components to consider in a particular cluster is dictated by the underlying geometry captured by the cluster and the strength of the corrupting noise. Once a dictionary is defined for a cluster, each patch in the cluster is denoised by expressing it as a linear combination of the dictionary elements. The coefficients of such a linear combination for any particular patch is determined in a regression framework using the local dictionary for the cluster. Each step of our method is well motivated and is shown to minimize some cost function. We then present an iterative extension of our algorithm that results in further performance gain. We validate our method through experiments with simulated as well as real noisy images. These indicate that our method is able to produce results that are quantitatively and qualitatively comparable to those obtained by some of the recently proposed state of the art denoising techniques.

Image Denoising via Improved Simultaneous Sparse Coding with Laplacian Scale Mixture

Wuhan University Journal of Natural Sciences, 2018

Image denoising is a well-studied problem closely related to sparse coding. Noticing that the Laplacian distribution has a strong sparseness, we use Laplacian scale mixture to model sparse coefficients. With the observation that prior information of an image is relevant to the estimation of sparse coefficients, we introduce the prior information into maximum a posteriori (MAP) estimation of sparse coefficients by an appropriate estimate of the probability density function. Extending to structured sparsity, a nonlocal image denoising model: Improved Simultaneous Sparse Coding with Laplacian Scale Mixture (ISSC-LSM) is proposed. The centering preprocessing, which admits biased-mean of sparse coefficients and saves expensive computation, is done firstly. By alternating minimization and learning an orthogonal PCA dictionary, an efficient algorithm with closed-form solutions is proposed. When applied to noise removal, our proposed ISSC-LSM can capture structured image features, and the adoption of image prior information leads to highly competitive denoising performance. Experimental results show that the proposed method often provides higher subjective and objective qualities than other competing approaches. Our method is most suitable for processing images with abundant self-repeating patterns by effectively suppressing undesirable artifacts while maintaining the textures and edges.

MAP-based image denoising with structured sparsity and Gaussian scale mixture

Pattern Analysis and Applications, 2018

Image denoising is a classical problem in image processing and is known to be closely related to sparse coding. In this work, based on the key observation that the probability density function (PDF) of image patch is relevant to the maximum a posteriori estimation of sparse coefficients, using an efficient approximation of the PDF of image patch, a nonlocal image denoising method: improved simultaneous sparse coding with Gaussian scale mixture (ISSC-GSM) is proposed. The preprocessing of centering for a collection of similar patches saves expensive computation and admits biased-mean of sparse coefficients. Our formulation can be efficiently computed by alternating minimization, and both subproblems have analytical solutions using the orthogonal PCA dictionary. When applied to noise removal, the proposed ISSC-GSM has achieved highly competitive denoising performance with often higher subjective and objective qualities than other competing approaches. Experimental results have shown that our method often provides the best visual quality by effectively suppressing undesirable artifacts while maintaining the textures and edges, which is most suitable for processing images with abundant self-repeating patterns.

Enhancing Image Denoising by Controlling Noise Incursion in Learned Dictionaries

IEEE Signal Processing Letters, 2015

Existing image denoising frameworks via sparse representation using learned dictionaries have an weakness that the dictionary, trained from noisy image, suffers from noise incursion. This paper analyzes this noise incursion, explicitly derives the noise component in the dictionary update step, and provides a simple remedy for a desired signal to noise ratio. The remedy is shown to perform better both in objective and subjective measures for lesser computation, and complements the framework of image denoising.

Denoising by low-rank and sparse representations

Journal of Visual Communication and Image Representation, 2016

Due to the ill-posed nature of image denoising problem, good image priors are of great importance for an effective restoration. Nonlocal self-similarity and sparsity are two popular and widely used image priors which have led to several state-of-the-art methods in natural image denoising. In this paper, we take advantage of these priors and propose a new denoising algorithm based on sparse and low-rank representation of image patches under a nonlocal framework. This framework consists of two complementary steps. In the first step, noise removal from groups of matched image patches is formulated as recovery of low-rank matrices from noisy data. This problem is then efficiently solved under asymptotic matrix reconstruction model based on recent results from random matrix theory which leads to a parameterfree optimal estimator. Nonlocal learned sparse representation is adopted in the second step to suppress artifacts introduced in the previous estimate. Experimental results, demonstrate the superior denoising performance of the proposed algorithm as compared with the state-of-the-art methods.

Real-world Noisy Image Denoising: A New Benchmark

Most of previous image denoising methods focus on additive white Gaussian noise (AWGN). However,the real-world noisy image denoising problem with the advancing of the computer vision techiniques. In order to promote the study on this problem while implementing the concurrent real-world image denoising datasets, we construct a new benchmark dataset which contains comprehensive real-world noisy images of different natural scenes. These images are captured by different cameras under different camera settings. We evaluate the different denoising methods on our new dataset as well as previous datasets. Extensive experimental results demonstrate that the recently proposed methods designed specifically for realistic noise removal based on sparse or low rank theories achieve better denoising performance and are more robust than other competing methods, and the newly proposed dataset is more challenging. The constructed dataset of real photographs is publicly available at https://github.com/csjunxu/PolyUDataset for researchers to investigate new real-world image denoising methods. We will add more analysis on the noise statistics in the real photographs of our new dataset in the next version of this article.

A Sparse Filtering-Based Approach for Non-blind Deep Image Denoising

2019

During the image acquisition process, some level of noise is usually added to the data mainly due to physical limitations of the sensor, and also regarding imprecisions during the data transmission and manipulation. Therefore, the resultant image needs to be further processed for noise attenuation without losing details. In this work, we attempt to denoise images using the advantage of sparse-based encoding and deep networks. Experiments on public images corrupted by different levels of Gaussian noise support the effectiveness of the proposed approach concerning some state-of-the-art image denoising approaches.