Showing papers on "Impulse noise published in 2016"
TL;DR: A nonconvex low-rank matrix approximation model is introduced that falls into the applicable scope of an augmented Lagrangian method, and its WSN minimization subproblem can be efficiently solved by generalized iterated shrinkage algorithm.
Abstract: Hyperspectral images (HSIs) are inevitably corrupted by mixture noise during their acquisition process, in which various kinds of noise, e.g., Gaussian noise, impulse noise, dead lines, and stripes, may exist concurrently. In this paper, mixture noise removal is well illustrated by the task of recovering the low-rank and sparse components of a given matrix, which is constructed by stacking vectorized HSI patches from all the bands at the same position. Instead of applying a traditional nuclear norm, a nonconvex low-rank regularizer, i.e., weighted Schatten p -norm (WSN), is introduced to not only give better approximation to the original low-rank assumption but also to consider the importance of different rank components. The resulted nonconvex low-rank matrix approximation (LRMA) model falls into the applicable scope of an augmented Lagrangian method, and its WSN minimization subproblem can be efficiently solved by generalized iterated shrinkage algorithm. Moreover, the proposed model is integrated into an iterative regularization schema to produce final results, leading to a completed HSI restoration framework. Extensive experimental testing on simulated and real data shows, both qualitatively and quantitatively, that the proposed method has achieved highly competent objective performance compared with several state-of-the-art HSI restoration methods.
134 citations
TL;DR: The proposed R NMF model is able to simultaneously handle Gaussian noise and sparse noise, and can be efficiently learned with elegant update rules, and sparsity regularizers are added to restrict the abundance maps in the RNMF, with the consideration of the sparse property of the material types within the hyperspectral scene.
Abstract: Hyperspectral unmixing (HU) is one of the crucial steps for many hyperspectral applications, including material classification and recognition. In the last decade, non-negative matrix factorization (NMF) and its extensions have been widely studied and have achieved advanced performances in HU. Unfortunately, most of the existing NMF-based methods make the assumption that the hyperspectral data are only corrupted by Gaussian noise. In real applications, the hyperspectral data are inevitably corrupted by sparse noise, which includes impulse noise, stripes, deadlines, and others types of noise. By separately modeling the sparse noise and Gaussian noise, a robust NMF (RNMF) model is subsequently introduced to unmix the hyperspectral data. The proposed RNMF model is able to simultaneously handle Gaussian noise and sparse noise, and can be efficiently learned with elegant update rules. In addition, sparsity regularizers are added to restrict the abundance maps in the RNMF, with the consideration of the sparse property of the material types within the hyperspectral scene. The experimental results with simulated and real data confirm the superiority of the proposed sparsity-regularized RNMF methods compared to the traditional NMF methods.
100 citations
TL;DR: Support vector machine (SVM) classification based Fuzzy filter (FF) is proposed for removal of impulse noise from gray scale images and suggests that this system outperforms some of the state of art methods while preserving structural similarity to a large extent.
84 citations
TL;DR: Simulation results indicate that the proposed method provides significant improvement over comparison filters especially for high noise densities.
46 citations
01 Sep 2016
TL;DR: The objective analysis suggests that there is ~3dB improvement in PSNR as compared to the MHFC based method for high density of impulse noise, and the results of SSIM along with visual observations indicate that the image details are maintained significantly in the proposed technique asCompared to existing methods.
Abstract: This paper proposes a multiclass support vector machine (SVM) based adaptive filter for removal of impulse noise from color images. The quality of the image gets degraded due to the presence of impulse noise. As a result, the homogeneity amongst the pixels gets distorted that needs to be restored. The feature set comprising of prediction error, difference between the median value and the center pixel; the median value in the kernel under operation has been used during this study. The pixel of test image is processed using adaptive window based filter that depends on the associated class assigned at the testing phase. The baseline system has been designed using modified histogram based fuzzy color filter (MHFC) technique. Four set of experiments have been carried out on a large database to validate the proposed method. The performance of the technique have been evaluated using peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM). The results suggest that for fixed valued impulse noise, the proposed filter performs better than the MHFC in case of high density impulse noise (>45%). However, for random valued impulse noise the proposed filter outperforms the MHFC based method for both low and high density of noise. The objective analysis suggests that there is ~3dB improvement in PSNR as compared to the MHFC based method for high density of impulse noise. The results of SSIM along with visual observations indicate that the image details are maintained significantly in the proposed technique as compared to existing methods.
41 citations
11 Jul 2016
TL;DR: Insight is provided into the suitability of the convolutional form for this type of application by comparing its performance as an image model with that of the standard model in an impulse noise restoration problem.
Abstract: Standard sparse representations, applied independently to a set of overlapping image blocks, are a very effective approach to a wide variety of image reconstruction problems. Convolutional sparse representations, which provide a single-valued representation optimised over an entire image, provide an alternative form of sparse representation that has recently started to attract interest for image reconstruction problems. The present paper provides some insight into the suitability of the convolutional form for this type of application by comparing its performance as an image model with that of the standard model in an impulse noise restoration problem.
34 citations
TL;DR: The proposed novel harmonic suppression method based on fractional lower order statistics (FLOS) has a competitive advantage that it can suppress harmonics well even if the impulse noise activating and has a fast tracking ability for changing harmonics.
Abstract: Impulse noise in power systems would seriously degrade the harmonic suppression performance. To remedy this problem, a novel harmonic suppression method based on fractional lower order statistics (FLOS) is proposed in this paper. In the proposed method, impulse noise is modeled by alpha-stable distribution. Then, the ESPRIT spectrum estimation algorithm is improved by FLOS for impulse noise and used to estimate the fundamental frequency of power signal, and the frequency of each harmonic component is obtained from this estimated frequency. Next, the amplitude of each harmonic component is estimated by a modified recursive least squares (RLS) algorithm. Finally, a harmonic compensation signal is generated by the active power filter based on the estimated frequencies and amplitudes to cancel original harmonics. The proposed method has a competitive advantage that it can suppress harmonics well even if the impulse noise activating and has a fast tracking ability for changing harmonics. Also, due to the use of self-sensing actuator principle, the proposed method can not only guarantee the performance of suppressing harmonics at normal operation states, but also ensure not to amplify harmonics in case of malfunction. The simulation results show that the proposed method has a better harmonic suppression performance than the existing ones under the impulse noise environment. The real experiments are also presented to verify the feasibility of the proposed method.
31 citations
TL;DR: The proposed solution is a quaternion switching vector filter in which the impulse detection consists of two stages and the concept of peer group is modified and extended to the directional samples to further detect whether they are corrupted by impulse noise or not.
31 citations
TL;DR: Simulation results prove that the PDBF has outperformed recently proposed state-of-the-art filters in terms of peak signal to noise ratio (PSNR), structural similarity index (SSIM), image enhancement factor (IEF), mean absolute error (MAE) and visual representation at the noise densities (ND) as high as 95%.
Abstract: A new probabilistic decision based filter (PDBF) is presented to remove salt and pepper impulse noise in highly corrupted images. The filter employs two types of estimation techniques for denoising namely trimmed median (TM) and patch else trimmed median (PETM) which is our main contribution in this paper. Depending upon the estimated noise density, the filter utilizes either TM or PETM and hence enhanced outcome of denoising. Simulation results prove that the PDBF has outperformed recently proposed state-of-the-art filters in terms of peak signal to noise ratio (PSNR), structural similarity index (SSIM), image enhancement factor (IEF), mean absolute error (MAE) and visual representation at the noise densities (ND) as high as 95%.
28 citations
TL;DR: The proposed filter uses the intensity based directional statistics to construct adaptive fuzzy membership functions which plays an important role in fuzzy inference system for random valued impulse noise detection and removal.
Abstract: Noise filtering in presence of important image detail information is considered as challenging task in imaging applications. Use of fuzzy logic based techniques is capturing more focus since last decade to deal with these challenges. In order to tackle conflicting issues of noise smoothing and detail preservation, this paper presents a novel approach using adaptive fuzzy inference system for random valued impulse noise detection and removal. The proposed filter uses the intensity based directional statistics to construct adaptive fuzzy membership functions which plays an important role in fuzzy inference system. Fuzzy inference system constructed in this way is used by the noise detector for accurate classification of noisy and noise-free pixels by differentiating them from edges and detailed information present in an image. After classification of pixels, noise adaptive filtering is performed based on median and directional median filter using the information provided by the noise detector. Simulation results based on well known quantitative measure i.e., peak-signal-to-noise ratio (PSNR) show the effectiveness of proposed filter.
26 citations
TL;DR: A proximal alternating direction method of multipliers (ADMM) is presented to solve the corrected TVL1 model and its convergence is also established under very mild conditions, and a comparison with a state-of-the-art method, the two-phase method, demonstrates the superiority of the proposed approach.
Abstract: For the problem of image restoration of observed images corrupted by blur and impulse noise, the widely used TVL1 model may deviate from both the data-acquisition model and the prior model, especially for high noise levels. In order to seek a solution of high recovery quality beyond the reach of the TVL1 model, we propose an adaptive correction procedure for TVL1 image deblurring under impulse noise. Then, a proximal alternating direction method of multipliers (ADMM) is presented to solve the corrected TVL1 model and its convergence is also established under very mild conditions. It is verified by numerical experiments that our proposed approach outperforms the TVL1 model in terms of signal-to-noise ratio (SNR) values and visual quality, especially for high noise levels: it can handle salt-and-pepper noise as high as 90% and random-valued noise as high as 70%. In addition, a comparison with a state-of-the-art method, the two-phase method, demonstrates the superiority of the proposed approach.
20 Nov 2016
TL;DR: A novel blind image denoising method under the Bayesian learning framework, which automatically performs noise inference and reconstructs the latent clean image by utilizing the patch group (PG) based image nonlocal self-similarity prior.
Abstract: Most existing image denoising methods assume to know the noise distributions, e.g., Gaussian noise, impulse noise, etc. However, in practice the noise distribution is usually unknown and is more complex, making image denoising still a challenging problem. In this paper, we propose a novel blind image denoising method under the Bayesian learning framework, which automatically performs noise inference and reconstructs the latent clean image. By utilizing the patch group (PG) based image nonlocal self-similarity prior, we model the PG variations as Mixture of Gaussians, whose parameters, including the number of components, are automatically inferred by variational Bayesian method. We then employ nonparametric Bayesian dictionary learning to extract the latent clean structures from the PG variations. The dictionaries and coefficients are automatically inferred by Gibbs sampling. The proposed method is evaluated on images with Gaussian noise, images with mixed Gaussian and impulse noise, and real noisy photographed images, in comparison with state-of-the-art denoising methods. Experimental results show that our proposed method performs consistently well on all types of noisy images in terms of both quantitative measure and visual quality, while those competing methods can only work well on the specific type of noisy images they are designed for and perform poorly on other types of noisy images. The proposed method provides a good solution to blind image denoising.
TL;DR: The new AHAAH model with ICE as the dose metric is adequate for use as a medical standard against impulse noise injury and the transfer functions from the new model are in good agreement with those of the human ear.
TL;DR: Auditory 4.0 is the only model known to date that provides the full TTS and PTS dose-response curves, including a TTS recovery model, and shows good agreement with historical data.
Abstract: Objectives: The new Auditory 4.0 model has been developed for the assessment of auditory outcomes, expressed as temporary threshold shift (TTS) and permanent threshold shift (PTS), from exposures to impulse noise for unprotected ears, including the prediction of TTS recovery. Methods: Auditory 4.0 is an empirical model, constructed from test data collected from chinchillas exposed to impulse noise in the laboratory. Injury outcomes are defined as TTS and PTS, and Auditory 4.0 provides the full range of TTS and PTS dose–response curves with the risk factor constructed from A-weighted sound exposure level. Human data from large weapons noise exposure was also used to guide the development of the recovery model. Results: Guided by data, a 28-dBA shift was applied to the dose–response curves to account for the scaling from chinchillas to humans. Historical data from rifle noise tests were used to validate the dose–response curves. New chinchilla tests were performed to collect recovery data to constr...
TL;DR: A new denoising method is introduced to remove noise in digital images by modeling this vagueness as entropy by utilizing interval-valued intuitionistic fuzzy set (IVIFS).
TL;DR: This work proposes a technique to remove sparse impulse noise from hyperspectral images by empirically learns the spatial and spectral sparsifying dictionaries while denoising the images based on the recently introduced Blind Compressed Sensing framework.
TL;DR: An incremental least-mean M-estimate algorithm is proposed for distributed network, which minimises a robust M-ESTimator-based cost function against impulsive noise via the gradient descent method.
Abstract: An incremental least-mean M-estimate algorithm is proposed for distributed network, which minimises a robust M-estimator-based cost function against impulsive noise via the gradient descent method. Moreover, based on the same frame for implementation, an incremental distributed least-mean p-power algorithm is presented for combating impulsive noise. Simulations verify the superiority of the proposed algorithms in the presence of impulsive noise relative to some existing incremental distributed algorithms.
TL;DR: A new image denoising method for impulse noise in greyscale images using a context-based prediction scheme is presented, which preserves the details in the filtered images better than other methods.
Abstract: A new image denoising method for impulse noise in greyscale images using a context-based prediction scheme is presented. The algorithm replaces the noisy pixel with the value occurring with the highest frequency, in the same context as the replaceable pixel. Since it is a context-based technique, it preserves the details in the filtered images better than other methods. In the aim of validation, the authors have compared the proposed method with several existing denoising methods, many of them being outperformed by the proposed filter.
TL;DR: Numerical experiments show that the proposed frame-based iterative algorithm compares favorably or often outperforms three well-known recent image-restoration methods employed for removing the mixed Gaussian and impulse noise.
Abstract: In this paper, we propose a frame-based iterative algorithm to restore images which are corrupted by mixed Gaussian and impulse noise, under the assumption that the image region corrupted by impulse noise is unknown. The removal of mixed Gaussian and impulse noise by our proposed algorithm is split into two subproblems which are solved alternatively and iteratively. With an initial guessed region of location for impulse noise, the first subproblem is to inpaint a corrupted image by solving a frame-based convex minimization scheme using the balanced approach, where sparse and redundant directional representations play a key role. Motivated by our recent work on frame-based image denoising and image inpainting, we shall employ the tight frame generated from the directional tensor product complex tight framelets in our balanced approach to remove the mixed Gaussian and impulse noise. Such tensor product complex tight framelets provide sparse directional representations for natural images and can capture the cartoon and texture parts of images very well. The second subproblem is to estimate the image region of locations where the pixels are corrupted by impulse noise. We solve the second subproblem using an $$l_0$$l0-minimization scheme. We consider both salt-and-pepper impulse noise and random-valued impulse noise. Numerical experiments show that our proposed algorithm compares favorably or often outperforms three well-known recent image-restoration methods employed for removing the mixed Gaussian and impulse noise.
23 Mar 2016
TL;DR: This paper proposes a modified decision based median filter that removes impulse noise from gray images that follows DBMF that considers only the noisy pixels and replaces the pixel value with median value of the pixels present in the processing window.
Abstract: Removing impulse noise in digital images is one of the major challenges in digital image processing. Pixels in digital images get corrupted during transmission due to impulse noise. In this paper, we propose a modified decision based median filter that removes impulse noise from gray images. For noise removal from digital images, different types of median filters are used: Standard Median Filter (MF), Weighted Median Filter (WMF), Adaptive Median Filter (AMF) and Decision Based Median Filter (DBMF). In most of these methods except DBMF, processing pixels, irrespective of the fact whether it is corrupted or not, are replaced by the median value of the pixels in their nearby region without considering the local features present for example edges. However, our proposed method follows DBMF that considers only the noisy pixels and replaces the pixel value with median value of the pixels present in the processing window. In our method, we increase the window size as per the requirement. Our experimental results show that our proposed method performs better than Standard Median Filter (MF), Weighted Median Filter (WMF), Adaptive Median Filter (AMF) and Decision Based Median Filter (DBMF), especially when the noise intensity level is high. We compare our method to others based on Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index (SSIM) values.
TL;DR: The results of the experiment show that the proposed algorithm successfully combines the merits of the Wiener filter and sharpening and achieves a significant proficiency in the enhancement of degraded X-ray images exhibiting Poisson noise, blurriness, and edge details.
Abstract: To resolve the problems of Poisson/impulse noise, blurriness, and sharpness in degraded X-ray images, a novel and efficient enhancement algorithm based on X-ray image fusion using a discrete wavelet transform is proposed in this paper. The proposed algorithm consists of two basics. First, it applies the techniques of boundary division to detect Poisson and impulse noise corrupted pixels and then uses the Wiener filter approach to restore those corrupted pixels. Second, it applies the sharpening technique to the same degraded X-ray image. Thus, it has two source X-ray images, which individually preserve the enhancement effects. The details and approximations of these sources X-ray images are fused via different fusion rules in the wavelet domain. The results of the experiment show that the proposed algorithm successfully combines the merits of the Wiener filter and sharpening and achieves a significant proficiency in the enhancement of degraded X-ray images exhibiting Poisson noise, blurriness, and edge details.
TL;DR: A new impulse noise detection algorithm is presented that is based on Noise ratio Estimation and a combination of K-means clustering and Non-Local Means based filter (NEK-NLM).
TL;DR: In this article, a novel algorithm for optical character recognition in the presence of impulse noise by applying a wavelet transform, principal component analysis, and neural networks was proposed, which can effectively recognize the characters in images in images.
Abstract: In this paper we propose a novel algorithm for optical character recognition in the presence of impulse noise by applying a wavelet transform, principal component analysis, and neural networks In the proposed algorithm, the Haar wavelet transform is used for low frequency components allocation, noise elimination and feature extraction The principal component analysis is used to reduce the dimension of the extracted features We use a set of different multi-layer neural networks as classifiers for each character; the inputs are represented by a reduced set of features One of the key features of the proposed approach is creating a separate neural network for each type of character The experimental results show that the proposed algorithm can effectively recognize the characters in images in the presence of impulse noise; the results are comparable with ABBYY FineReader and Tesseract OCR
TL;DR: This work proposes an image denoising algorithm through skillfully combining NLM and sparse representation technique to remove Gaussian noise mixed with random-valued impulse noise and can reconstruct the clean image.
TL;DR: Simulation results illustrate that receivers with combined turbo coding and the proposed noise compensator drastically outperform existing receivers under impulsive noise.
Abstract: In this paper, we propose a novel estimation and decoding scheme for power-line communication (PLC) systems in impulsive noise environments. The proposed scheme is based on the turbo coding combined with adaptive noise compensation to reduce burst errors and multipath effects. For this purpose, the PLC channel and noise models are introduced, then, the turbo encoder/decoder are inserted in the mapper/demapper and the pilot insertion block for the sake of enhancing preliminary estimation of the transmitted orthogonal frequency division multiplexing signals. The proposed impulsive noise compensator is based on the estimation of the impulse bursts using a new blanking/clipping function, and on the estimation of the signal to impulse noise ratio and the peak to average power ratio. Simulation results illustrate that receivers with combined turbo coding and the proposed noise compensator drastically outperform existing receivers under impulsive noise. In comparison to some existing schemes, the proposed scheme reaches its perfect performance in a reduced 15-paths environment, when the bit error rate and mean square error performance are tested. The improvements in SNR performance are more than 16 dB for BPSK modulation and can reach 12---20 dB for 16-QAM modulation, when a high impulsive noise level is considered.
TL;DR: A genetic expression programming-based classifier is employed for the detection of impulse noise-corrupted pixels and a modified median filter is used to reduce the blurring effect caused due to filtering operation on the noise-free pixels.
Abstract: Existing impulse noise reduction techniques perform well at low noise densities; however, their performance drops sharply at higher noise densities. In this paper, we propose a two-stage scheme to surmount this problem. In the proposed approach, first stage consists of impulse detection unit followed by the filtering operation in the second stage. We have employed a genetic expression programming-based classifier for the detection of impulse noise-corrupted pixels. To reduce the blurring effect caused due to filtering operation on the noise-free pixels, we filter the detected noisy pixels only by using a modified median filter. Better peak signal-to-noise ratio, structural similarity index measure, and visual output imply the efficacy of the proposed scheme for noise reduction at higher noise densities.
TL;DR: Experimental studies show that the proposed technique is simple, easy to implement, robust to noise, and outperforms the classic vector filters, as well as more recent filters.
Abstract: This research presents a complete study of a new alternating vector filter for the removal of impulsive noise in colour images. The method is based on an impulsive noise detector for greyscale images that has been adapted in a localised manner using geometric information for processing colour images. Based on this statistic, a filtering scheme alternating between the identity and a non-linear vector filter is proposed. A geometric and experimental study was performed to obtain the optimal filter design. Experimental studies show that the proposed technique is simple, easy to implement, robust to noise, and outperforms the classic vector filters, as well as more recent filters.
TL;DR: This paper presents an implementation of the PARIGI method, which relies on a patch-based approach, which requires careful choices for both the distance between patches and for the statistical estimator of the original patch.
Abstract: In this paper, we present an implementation of the PARIGI method that addresses the problem of the restoration of images affected by impulse noise or by a mixture of Gaussian and impulse noise. The method relies on a patch-based approach, which requires careful choices for both the distance between patches and for the statistical estimator of the original patch. Experiments are performed in the case of pure impulse noise and in the case of a mixture of Gaussian and impulse noise.
TL;DR: Experimental results show that iterative impulse noise filters with the proposed automatic filtering convergence method can remove much of the impulse noise and effectively maintain image details and operate more efficiently.
Abstract: We proposed an automatic method to improve performance of iterative noise filters.The iterative noise filters with the automatic method can process in finite steps.This method showed better implementation for de-noising in experimental results. Whether input images are corrupted by impulse noise and what the noise density level is are unknown a priori, and thus published iterative impulse noise filters cannot adaptively reduce noise, resulting in a smoothing image or unclear de-noising. For this reason, this paper proposes an automatic filtering convergence method using PSNR checking and filtered pixel detection for iterative impulse noise filters. (1) First, the similarity between the input image and the 1st filtered image is determined by calculating MSE. If MSE is equal to 0, then the input image is unfiltered and becomes the output. (2) Otherwise, one applies PSNR checking and filtered pixel detection to estimate the difference between the tth filtered image and the t-1th filtered image. (3) Finally, an adaptive and reasonable threshold is defined to make the iterative impulse noise filters stop automatically for most image details preservation in finite steps. Experimental results show that iterative impulse noise filters with the proposed automatic filtering convergence method can remove much of the impulse noise and effectively maintain image details. In addition, iterative impulse noise filters operate more efficiently.
TL;DR: In this article, a variational multiphase image segmentation model based on fuzzy membership functions and L1-norm fidelity is proposed, which is more robust to outliers such as impulse noise and keeps better contrast.
Abstract: In this paper, we propose a variational multiphase image segmentation model based on fuzzy membership functions and L1-norm fidelity. Then we apply the alternating direction method of multipliers to solve an equivalent problem. All the subproblems can be solved efficiently. Specifically, we propose a fast method to calculate the fuzzy median. Experimental results and comparisons show that the L1-norm based method is more robust to outliers such as impulse noise and keeps better contrast than its L2-norm counterpart. Theoretically, we prove the existence of the minimizer and analyze the convergence of the algorithm.