De-noising Effect Research Articles

Multivariate Curve Resolution for 2D Solid-State NMR spectra.

We present a processing method, based on the multivariate curve resolution approach (MCR), to denoise 2D solid-state NMR spectra, yielding a substantial S/N ratio increase while preserving the lineshapes and relative signal intensities. These spectral features are particularly important in the quantification of silicon species, where sensitivity is limited by the low natural abundance of the 29Si nuclei and by the dilution of the intrinsic protons of silica, but can be of interest also when dealing with other intermediate-to-low receptivity nuclei. This method also offers the possibility of coprocessing multiple 2D spectra that have the signals at the same frequencies but with different intensities (e.g.: as a result of a variation in the mixing time). The processing can be carried out on the time-domain data, thus preserving the possibility of applying further processing to the data. As a demonstration, we have applied Cadzow denoising on the MCR-processed FIDs, achieving a further increase in the S/N ratio and more effective denoising also on the transients at longer indirect evolution times. We have applied the combined denoising on a set of experimental data from a lysozyme–silica composite.

Channel Estimation Based on Adaptive Denoising for Underwater Acoustic OFDM Systems

Underwater acoustic (UWA) communications systems suffer from a low signal-to-noise ratio (SNR) and a doubly selective channel, which are caused by many limitations, such as high propagation loss, slow propagation speed, and time-varying environmental factors. To overcome a low SNR, various diversity techniques are often adopted in UWA communications. However, such diversity can only be exploited with knowledge of the channel. Accordingly, channel estimation needs to be performed without obtaining the benefit of diversity. Moreover, accurate side information that can support a channel estimator (CE) is difficult to acquire at a low SNR under doubly selective channel. In this article, a novel CE based on an adaptive denoising is proposed for UWA orthogonal frequency-division multiplexing (OFDM) systems. The proposed method exploits two different types of pilot symbols. Channel impulse response (CIR) is estimated based on primary pilot symbols. By minimizing the squared error of the received secondary pilot symbols, a near-optimal denoising window is adaptively determined based on the channel length for the given CIR estimate. The proposed method does not require a priori information about channel statistics and SNR values. Analysis on the effect of denoising and the performance of the proposed denoising window estimator are also presented. Simulation and at-sea experiments verify that the proposed method has superior performance, compared with conventional CEs over diverse channel conditions. Complexity analysis shows that the proposed method is computationally efficient. Therefore, the proposed method is effective for real-time UWA OFDM systems under a harsh UWA channel with strong noise.

An Improved Combination of Image Denoisers Using Spatial Local Fusion Strategy

Image denoising is a well-researched problem in the image processing field. Numerous image denoising algorithms have been proposed in the past. Although researchers have continually focused on improving the denoising algorithm performance regarding denoising effect and outstanding results have been achieved, the improvement amplitude of a single denoising algorithm has decreased over time. Recently, the consensus neural network (CsNet), which combines multiple image denoisers to produce an overall better result compared to single algorithms, was proposed. However, the denoising process of CsNet is time-consuming owing to its MSE optimal weight setting and iterative boosting stages. Therefore, we propose an improved combination of nonlocally centralized sparse representation (NCSR) with a fast and flexible denoising convolutional neural network (FFDNet) using a spatial local fusion strategy (ICID). ICID uses a structural-based patch to decompose their denoised images into the strength, structure, and mean intensity components. Thereafter, an image patch is reconstructed and placed back into the fused image after fusing the three components separately. Experimental results verified that our algorithm is superior to CsNet, and it is faster. The combination of NCSR and FFDNet can harmonize the complementary denoising capabilities of different denoising algorithms. NCSR can preserve as many details as possible in natural images with numerous repeated structures, whereas FFDNet can achieve state-of-the-art results with a sufficiently large training set of images. Moreover, ICID uses the structural-based method, which considers more local details and preserves more textures, resulting in superior performance.

Noise Reduction on Received Signals in Wireless Ultraviolet Communications Using Wavelet Transform

In wireless ultraviolet optical communications, noise is one of the most essential factors affecting the communication system performance. This paper presents a scheme of reducing noise in the received ultraviolet signal using wavelet transform algorithms. An effective signal-to-noise ratio (SNR) calculation method for the received signals is also proposed, and adopted by our wavelet denoising scheme so that an optimal wavelet basis function can be selected. The proposed denoising method is applied to both the ultraviolet signals generated from the multi-scattering transmission simulation model and the signals received from physical experiments, under different conditions of transceiver elevation angles and communication distances. The results show consistently that the wavelet transform algorithm can significantly improve the SNRs at the receiving end. When the wavelet basis is coif2, the best denoising effect is achieved where the improved SNR reaches 11.9925 dB on average for various physical conditions.

Parallel mining method for symbol application features of complex network images

Aiming at the problems of poor denoising effect, low recognition rate and long feature mining time in current methods, a parallel feature mining method of image symbol application features in complex network based on neural network learning control is proposed. First, the standard deviation of image symbol noise is calculated and denoising is carried out by filtering parameters. Then the image symbol is segmented by using the two-dimensional maximum inter-group variance method. Finally, the momentum coefficient and learning efficiency are introduced to calculate the parameters of the neural network, and the improved simulated annealing algorithm is used to adjust the learning efficiency of the neural network, so as to realise the parallel mining of image symbol application features of the complex network. Experimental results show that this method has good image denoising effect, high image recognition rate and short feature mining time, which verifies the comprehensive effectiveness of this method.

Parallel mining method for symbol application features of complex network images

Aiming at the problems of poor denoising effect, low recognition rate and long feature mining time in current methods, a parallel feature mining method of image symbol application features in complex network based on neural network learning control is proposed. First, the standard deviation of image symbol noise is calculated and denoising is carried out by filtering parameters. Then the image symbol is segmented by using the two-dimensional maximum inter-group variance method. Finally, the momentum coefficient and learning efficiency are introduced to calculate the parameters of the neural network, and the improved simulated annealing algorithm is used to adjust the learning efficiency of the neural network, so as to realise the parallel mining of image symbol application features of the complex network. Experimental results show that this method has good image denoising effect, high image recognition rate and short feature mining time, which verifies the comprehensive effectiveness of this method.

Study on Denoise Algorithm for Micro-motion Target

Aiming at the non-stationary characteristics of narrowband radar echo of micro-motion target, a novel micro-motion signal denoising algorithm is proposed. Firstly, the weight function is constructed by coherent accumulation, and the echo is weighted by weight function to filter out the noise in range dimension. Secondly, the singular spectrum analysis(SSA) is performed on the range bin of the echo, the signal with large singularities is preserved, and the noise with small singularities is filtered out. Then, the azimuth dimension signal transformed from time domain into time-frequency domain, the step iteration method is used to adaptively filter the time-frequency spectrum and the time-frequency filtering result with the best denoising effect selected by defining the risk function. Finally, the simulation verify the effect of the proposed method.

Abnormal node detection method for time-sharing heating energy consumption in multi-storey buildings based on drosophila algorithm

In order to solve the problem of high energy consumption of time-sharing heating in multi-storey buildings caused by abnormal node, it is necessary to study the detection method of abnormal node of heating energy consumption. Based on fruit flies algorithm of multi-storey building time-sharing heating energy consumption of the abnormal node detection method, using synchronous sequential PLD multi-storey building heating system of heating time-sharing data, put optimisation algorithm is introduced into EEMD algorithm, noise by flies optimisation calculation the corresponding threshold, remove the noise, the IMF component complete data denoising processing. Based on the recursive dynamic principal component analysis method, the detection model of abnormal heating energy consumption nodes is constructed to realise the detection of abnormal heating energy consumption nodes in multi-storey buildings. Experimental results show that the proposed method has good denoising effect, low false alarm rate and high detection rate.

An effective gamma white spots removal method for CCD-based neutron images denoising

In CCD-based neutron radiography and tomography, resulting digital neutron images are usually degraded by gamma white spots caused by the interaction between high energy gamma radiation and the charge-coupled device (CCD) detector. The gamma white spots have a major impact on the accuracy of quantitative data analysis and are difficult to remove due to their high magnitudes and non-uniform distribution. In this paper, we put forward a novel iterative filtering method based on a “detection-corrosion” strategy for degraded neutron image denoising. A collaborative filtering method is adopted via integrating a spatially adaptive filter and a median filter for gamma white spots detection and removal. The experimental results show that the proposed method can achieve satisfactory denoising effect with preferable preservation of image details.

Image De-noising using Optimized Self Similar Patch based Filter

Emerging trends in the widespread use of technology has led to proliferation of images and videos acquired and distributed through electronic devices. There is an increasing need towards capturing high fidelity images and filtering of the concomitant noise inevitable in the capture, transmission and reception of the same. In this paper, we propose an OPSS (Optimized Patch based Self Similar) filter that exploits concurrently the photometric, geometric and graphical patch similarities of the image. This model recognizes similar patches to segregate the corrupted from the uncorrupted pixels in an image and improve the performance of denoising. Photometric patch similarity is established by using Non-Local Means Decision Based Unsymmetrical Trimmed Median (NLM-DBUTM) filter, which computes weights based on the reference patch. The geometrical patch similarity is done through the K-means clustering and graphically similar patches are identified through Ant Colony Optimization (ACO) technique. These “three similarities” based models have been taken advantage of and combined to arrive at a more comprehensive and effective denoising. The results obtained through the OPSS algorithm demonstrate improved efficiency in removing Gaussian and Impulse noise. Experimental results demonstrate that our proposed study achieves good performance with respect to other denoising algorithms being compared. Experimental results are based on performance measure (evaluation parameters) through Peak Signal to Noise Ratio (PSNR), Mean squared error (MSE) and Structural Similarity Index Measure (SSIM).

Separable property-based super-resolution of lousy image data

This paper presents a novel wavelet-based approach for single-image super-resolution. Our technique integrates wavelet transform and the learned locally regularized anchored neighborhood regression model for more robust frequency estimation and image restoration. First, we decomposed the low-resolution input image into four frequency sub-bands by applying discrete wavelet transform and then processed these frequency sub-bands based on separable property of neighborhood filtering to achieve a fast parallel and vectorized operation by reducing computational burden resulting from computing the weighted function of every pixel for each pixel in an image. We then applied inverse discrete wavelet transform to reconstruct the original image. Super-resolution is achieved using the learned model to predict the high-resolution image features. Lastly, we explicitly unified both the locality structure and nonlocal self-similarity properties in natural image and incorporated them into our super-resolution framework to regularize the nonlinear correlation between low-resolution and high-resolution space and improve the reconstructed results. Experiments on standard images validate the effectiveness of our proposed method for effective denoising, deblurring and super-resolution reconstruction tasks compared to other top performing state-of-the-art methods both quantitatively and qualitatively.

Application of SNMF-MSS in low-frequency seismic noise suppression in desert areas

Noise suppression is an important step in seismic data processing. Various geographical environments make random noise in the acquired seismic records different in characteristics. Random noise in desert areas has non-Gaussian, non-linearity, non-stationary and heterogeneity. What is more, its frequency range is generally below 15 Hz, showing distinct low-frequency characteristics. Therefore, it brings much difficulty to the suppression of desert noise. Most of the existing noise suppression methods perform well on the condition of high-frequency noise, but perform worse for low-frequency noise because the denoising process of some methods depends on the waveform of noisy signal. And in some transformed methods, signal and noise are mixed with each other in the transformed domain. Thus, to resolve this problem, this paper presents a novel multiband spectrum separation (MSS) method based upon sparse nonnegative matrix factorization (SNMF) algorithm. It firstly decomposes noisy signal into some sub-signals of different narrow frequencies bands. Then, signal and noise components of these sub-signals are further separated by spectral unmixing in time-frequency domain. Several groups of experiments were carried out on synthetic records and real seismic data in Tarim oilfield in the northwest of China. Moreover, we compare its denoising effect with time-frequency peak filtering (TFPF), empirical mode decomposing (EMD) and Radon transform methods. The experimental and comparison results demonstrate that the proposed SNMF-MSS method performs better than the other three methods in the aspect of low-frequency suppression and weak signal recovery.

Image Denoising Algorithm based on Improved NCSR Model

The Nonlocally Centralized Sparse Representation (NCSR) algorithm aims to reduce sparse coding noise and utilizes the Nonlocal Self-similarity of the image to improve the effectiveness and accuracy of the image denoising algorithm, which has achieved good denoising effect. However, in the local dictionary learning and sparse coefficient estimation, the algorithm directly uses the Euclidean distance of the noise image to measure the similarity, which will have a certain impact on the final denoising effect. Aiming at this deficiency, it is proposed to use the Gaussian Mixed Model (GMM) to train nonlocal self-similar image patch of the external clean image in the local dictionary learning, and use the learned GMM to guide the internal image clustering; when performing the sparse coefficient estimation, the structural similarity index(SSIM) is introduced into the Euclidean distance and the improved Euclidean distance is introduced into the weight calculation. The experimental results show that the improved algorithm proposed in this paper has improved the PSNR and the visual effect of the image.

Fractional‐integral‐operator‐based improved SVM for filtering salt‐and‐pepper noise

To protect edge and texture information, when removing salt-and-pepper (SP) noise in grayscale images, a support vector machine (SVM) denoising method is employed. First, a mapping relation between the neighborhood signal pixels and the central pixel is designed. The size of the neighborhood is a 5 × 5 region, with a signal pixel in the center. In this region, a 25-dimensional input sample is constructed using the correlation between the neighborhood pixels and the eight-direction fractional integral operators. The center signal pixel acts as the corresponding output sample to provide a training sample. Then, the SVM is trained with all training samples, and the SVM denoising model is obtained. Next, the center pixel value is estimated using the SVM denoising model in every 5 × 5 region with a noise pixel in the center. Finally, the noise pixel values are replaced with the estimated values of the SVM. The experiments demonstrate that the best denoising effect is obtained when the fractional integral order is in the range of 1.8 ± 0.1. The proposed method produces a visually pleasing denoised image and obtains superior image quality assessment indicators. Our method has significant advantages compared with state-of-the-art denoisers when a low level of noise is present.

Deep Learning Based Noise Reduction for Brain MR Imaging: Tests on Phantoms and Healthy Volunteers.

Purpose:To test whether our proposed denoising approach with deep learning-based reconstruction (dDLR) can effectively denoise brain MR images.Methods:In an initial experimental study, we obtained brain images from five volunteers and added different artificial noise levels. Denoising was applied to the modified images using a denoising convolutional neural network (DnCNN), a shrinkage convolutional neural network (SCNN), and dDLR. Using these brain MR images, we compared the structural similarity (SSIM) index and peak signal-to-noise ratio (PSNR) between the three denoising methods. Two neuroradiologists assessed the image quality of the three types of images. In the clinical study, we evaluated the denoising effect of dDLR in brain images with different levels of actual noise such as thermal noise. Specifically, we obtained 2D-T2-weighted image, 2D-fluid-attenuated inversion recovery (FLAIR) and 3D-magnetization-prepared rapid acquisition with gradient echo (MPRAGE) from 15 healthy volunteers at two different settings for the number of image acquisitions (NAQ): NAQ2 and NAQ5. We reconstructed dDLR-processed NAQ2 from NAQ2, then compared with SSIM and PSNR. Two neuroradiologists separately assessed the image quality of NAQ5, NAQ2 and dDLR-NAQ2. Statistical analysis was performed in the experimental and clinical study. In the clinical study, the inter-observer agreement was also assessed.Results:In the experimental study, PSNR and SSIM for dDLR were statistically higher than those of DnCNN and SCNN (P < 0.001). The image quality of dDLR was also superior to DnCNN and SCNN. In the clinical study, dDLR-NAQ2 was significantly better than NAQ2 images for SSIM and PSNR in all three sequences (P < 0.05), except for PSNR in FLAIR. For all qualitative items, dDLR-NAQ2 had equivalent or better image quality than NAQ5, and superior quality to that of NAQ2 (P < 0.05), for all criteria except artifact. The inter-observer agreement ranged from substantial to near perfect.Conclusion:dDLR reduces image noise while preserving image quality on brain MR images.

Dynamic Residual Dense Network for Image Denoising.

Deep convolutional neural networks have achieved great performance on various image restoration tasks. Specifically, the residual dense network (RDN) has achieved great results on image noise reduction by cascading multiple residual dense blocks (RDBs) to make full use of the hierarchical feature. However, the RDN only performs well in denoising on a single noise level, and the computational cost of the RDN increases significantly with the increase in the number of RDBs, and this only slightly improves the effect of denoising. To overcome this, we propose the dynamic residual dense network (DRDN), a dynamic network that can selectively skip some RDBs based on the noise amount of the input image. Moreover, the DRDN allows modifying the denoising strength to manually get the best outputs, which can make the network more effective for real-world denoising. Our proposed DRDN can perform better than the RDN and reduces the computational cost by . Furthermore, we surpass the state-of-the-art CBDNet by 1.34 dB on the real-world noise benchmark.

DETERMINATION OF ROLLER BEARING INNER RACE DEFECT BASED ON VIBRATION SIGNAL

The present article’s intent is to measure and identify the roller bearing inner race defect width and its corresponding characteristic frequency based on filtered time-domain vibration signal. In case localized fault occurs in a bearing, the rolling elements encounter some slippage as the rolling elements enter and leave the bearing load zone. As a consequence, the incidence of the impacts never reproduce exactly at the same position from one cycle to another. Moreover, when the position of the defect is moving with respect to the load distribution zone of the bearing, the series of impulses are modulated in amplitude in time-domain and the conforming Bearing Characteristic Frequencies (BCFs) arise in frequency domain. In order to verify the ability of time-domain in measuring the fault of rolling bearing, an artificial fault is introduced in the vehicle gearbox bearing: an orthogonal placed groove on the inner race with the initial width of 0.6mm approximately. The faulted bearing is a roller bearing quantification of the characteristic features relevant to the inner race bearing defect. It is located on the gearbox input shaft—on the clutch side. To jettison the frequency associated with interferential vibrations, the vibration signal is filtered with a band-pass filter based on an optimal daughter Morlet wavelet function whose parameters are optimized based on maximum Kurtosis (Kurt.). The residual signal is performed for the measurement of defect width. The proposed technique is used to analyse the experimental signal of vehicle gearbox rolling bearing. The experimental test stand is equipped with two dynamometer machines; the input dynamometer serves as an internal combustion engine, the output dynamometer introduces the load on the flange of the output joint shaft. The Kurtosis and Pulse Indicator (PI) are selected as the evaluation parameters of the de-noising effect. The results show the reliability of the proposed approach for identification and quantification of the characteristic features relevant to the inner race bearing defect.

Analysis of MFL Model for Sucker Rod Defects and Its MFL Signal Processing

Abstract Magnetic flux leakage (MFL) testing is a nondestructive testing technique widely used in the petroleum industry and pipeline transportation. In this article, the double-coil magnetization method was used to analyze and solve the MFL model of a sucker rod defect, and the collected MFL signals were processed and analyzed. The model of the sucker rod defect was established using the magnetic dipole model, and the leakage magnetic field intensity in the tangential and normal directions above the defect was solved. Then, the influences of the defect width, defect depth, and lift-off value on the leakage magnetic field were obtained. The magnetic field gradients of the MFL signals were deduced and discussed, and the trend of the magnetic field gradients in different directions were obtained. The cubic spline interpolation method (CSIM) was introduced and was applied to the interpolation of MFL data. Finally, the MFL testing platform of the sucker rod defect was constructed, and the collected MFL data were denoised by the wavelet filtering (WF) method; subsequently, the magnetic gradients of the MFL signals were analyzed. The experimental result shows that the continuous MFL signals can be achieved by using the CSIM, and the WF not only maintains the characteristic information of the effective signals but also has a good denoising effect and makes the signals smooth. Simultaneously, the result also indicates that the magnetic field gradients are good tools for analyzing the MFL signals, which reflect the characteristic information of the defect.

An novel image de-noising model based on gradient and adaptive curvature features and its application

In this paper, an image de-noising model with gradient and adaptive curvature features is proposed for the visual inspection of the appearance defects of high-density flexible integrated circuit package substrates with strict line-width and line distance. Firstly, the model proposed in this paper adaptively adjusts the weight of the level set curvature feature and the gradient feature of the image, and uses more abundant first-order differential and second-order differential information of the image as the detection factor for image de-noising. Secondly, theoretical analysis shows that the diffusion intensity of the model in the flat region is larger than that of the classical model, and the de-noising effect is better. Furthermore, at the corners and peaks of the image, the proposed model can suppress the reduction of the gray value and preserve more detailed information of the image and edges. Finally, the experimental analysis shows that the proposed model has the best de-noising effect compared with other models. The method proposed in this paper can effectively remove the noise of the image of the high-density flexible integrated circuit package substrate, and at the same time retain more original details and edge information of the image which has practical significance.

Chest X‐ray image denoising method based on deep convolution neural network

To improve the visual effect of chest X-ray images and reduce the noise interference in disease diagnosis based on the chest X-ray images, the authors proposed an image denoising model based on deep convolution neural network. They utilise batch normalisation to solve the problem of performance degradation due to the increase of neural network layers, and use residual learning of the distribution of noise in noisy X-ray images. Specifically, the depthwise separable convolution is used to accelerate the convergence speed of network model, shorten the training time, and improve accuracy of the model. Compared to the several popular or the state-of-the-art denoising algorithms, their extensive experiments demonstrate that their method can not only achieve better denoising effects, but also significantly reduce the complexity of the network and shorten the computation time.