De-noising Effect Research Articles

Deep High-Resolution Network for Low-Dose X-Ray CT Denoising

Low-dose computed tomography (LDCT) is clinically desirable because it reduces the radiation dose to patients. However, the quality of LDCT images is often suboptimal because of the inevitable strong quantum noise. Because of their unprecedented success in computer vision, deep learning (DL)-based techniques have been used for LDCT denoising. Despite DL models’ promising ability to remove noise, researchers have observed that the resolution of DL-denoised images is compromised, which decreases their clinical value. To mitigate this problem, in this work, we developed a more effective denoiser by introducing a high-resolution network (HRNet). HRNet consists of multiple branches of subnetworks that extract multiscale features that are fused together later, which substantially enhances the quality of the generated features and improves denoising performance. Experimental results demonstrated that the introduced HRNet-based denoiser outperformed the benchmarked U-Net–based denoiser, as it provided superior image resolution preservation and comparable, if not better, noise suppression. Quantitative evaluation in terms of root-mean-squared errors (RMSEs)/structure similarity index (SSIM) showed that the HRNet-based denoiser improve these values from 113.80/0.550 (LDCT) to 55.24/0.745 (HRNet), which outperformed the 59.87/0.712 for the U-Net–based denoiser.

EFFECT OF WAVELET DE-NOISING ON THE CLASSIFICATION OF PIG BEHAVIOUR

To efficiently eliminate the noise generated by the triaxial accelerometer when collecting pigs’ behavioural data, this paper adopted SNR and MSE as the indexes to evaluate the de-noising effect of pigs’ acceleration signal under various combinations of wavelet basis, decomposition layer, threshold rule and threshold function. Based on the optimal wavelet parameter combinations, the de-noised data were divided into a training dataset and test dataset to conduct a 3-fold cross validation. The results showed that Db4 wavelet can achieve a satisfactory de-noising effect when used as a wavelet basis for 8 layers wavelet decomposition based on Rigrsure threshold rules and the new improved threshold function. As a result, compared with traditional wavelet hard threshold de-noising, soft threshold de-noising and EMD de-noising method, the improved threshold function improved the stability of signal filtering, which was shown to be more practical, effective and feasible. As such, wavelet de-noising was found to significantly improve the classification accuracy of all four behaviour classes (lying, standing, walking and exploring) considered for this study, and the overall major mean accuracy was improved from 0.680 to 0.826.

Distributed Spectrum Management in Cognitive Radio Networks by Consensus-Based Reinforcement Learning.

In this paper, we propose a new algorithm for distributed spectrum sensing and channel selection in cognitive radio networks based on consensus. The algorithm operates within a multi-agent reinforcement learning scheme. The proposed consensus strategy, implemented over a directed, typically sparse, time-varying low-bandwidth communication network, enforces collaboration between the agents in a completely decentralized and distributed way. The motivation for the proposed approach comes directly from typical cognitive radio networks’ practical scenarios, where such a decentralized setting and distributed operation is of essential importance. Specifically, the proposed setting provides all the agents, in unknown environmental and application conditions, with viable network-wide information. Hence, a set of participating agents becomes capable of successful calculation of the optimal joint spectrum sensing and channel selection strategy even if the individual agents are not. The proposed algorithm is, by its nature, scalable and robust to node and link failures. The paper presents a detailed discussion and analysis of the algorithm’s characteristics, including the effects of denoising, the possibility of organizing coordinated actions, and the convergence rate improvement induced by the consensus scheme. The results of extensive simulations demonstrate the high effectiveness of the proposed algorithm, and that its behavior is close to the centralized scheme even in the case of sparse neighbor-based inter-node communication.

Denoising of LCR Wave Signal of Residual Stress for Rail Surface Based on Lifting Scheme Wavelet Packet Transform

According to the acousto elastic effect, the residual stress on the surface of the rail can be evaluated by measuring the change in the propagation velocity of ultrasonic waves, such as longitudinal critically refracted (LCR) waves on the surface of the rail. The LCR wave signal is often polluted by a variety of noise sources, coupled with the influence of the poor surface condition of the inspected component, which greatly reduces the detectability and online measurement ability of the LCR wave signal. This paper proposes the application of the lifting scheme wavelet packet transform (LSWPT) denoising method to solve the noise suppression problem of LCR wave signal. The traditional wavelet transform (WT), wavelet packet transform (WPT), as well as the lifting scheme wavelet transform (LSWT) and lifting scheme wavelet packet transform are compared and analyzed in the soft thresholding and hard thresholding processing of denoising ability and efficiency of the noisy LCR wave signal. The experimental results show that the LSWPT method has the characteristics of fast calculation speed and a good denoising effect, and it is an efficient method of denoising signals for on-line ultrasonic measurement of residual stress on the rail surface.

RETRACTED ARTICLE: Big Data Mining and Analysis Based on Convolutional Fuzzy Neural Network

The purpose of this paper is to solve the shortcomings of traditional big data mining and analysis (BDMA); this paper compares and analyzes the denoise effect of wavelet transform and wavelet packet analysis and chooses to use wavelet packet analysis method to denoise the signal and extract fault feature vector. Finally, the extracted feature vectors are divided into two libraries, including training and testing. The fuzzy convolutional neural network (FCNN) is trained by using the training sample database data, the network weights are constantly modified, and the projection of nonlinear is computed between the input features and output features. After the expected recognition accuracy is achieved, the performance of FCNN is evaluated by using the detection samples. The performance comparison and analysis are conducted with the traditional BDMA algorithm, mainly including the comparison of recognition accuracy and learning convergence speed. Two stages of feature extraction and data mining (DM) are constructed on the big data set. The experimental results prove the effectiveness of wavelet transform feature and FCNN in analyzing and mining in big data. Finally, through the example of BDMA, this paper verifies that the organic combination of wavelet transform feature extraction and FCNN algorithm obviously improves the efficiency, meets the requirements of big data analysis, and provides potential application value for BDMA.

Toward a priori noise characterization for real-time edge-aware denoising in fluoroscopic devices

BackgroundLow-dose X-ray images have become increasingly popular in the last decades, due to the need to guarantee the lowest reasonable patient’s exposure. Dose reduction causes a substantial increase of quantum noise, which needs to be suitably suppressed. In particular, real-time denoising is required to support common interventional fluoroscopy procedures. The knowledge of noise statistics provides precious information that helps to improve denoising performances, thus making noise estimation a crucial task for effective denoising strategies. Noise statistics depend on different factors, but are mainly influenced by the X-ray tube settings, which may vary even within the same procedure. This complicates real-time denoising, because noise estimation should be repeated after any changes in tube settings, which would be hardly feasible in practice. This work investigates the feasibility of an a priori characterization of noise for a single fluoroscopic device, which would obviate the need for inferring noise statics prior to each new images acquisition. The noise estimation algorithm used in this study was tested in silico to assess its accuracy and reliability. Then, real sequences were acquired by imaging two different X-ray phantoms via a commercial fluoroscopic device at various X-ray tube settings. Finally, noise estimation was performed to assess the matching of noise statistics inferred from two different sequences, acquired independently in the same operating conditions.ResultsThe noise estimation algorithm proved capable of retrieving noise statistics, regardless of the particular imaged scene, also achieving good results even by using only 10 frames (mean percentage error lower than 2%). The tests performed on the real fluoroscopic sequences confirmed that the estimated noise statistics are independent of the particular informational content of the scene from which they have been inferred, as they turned out to be consistent in sequences of the two different phantoms acquired independently with the same X-ray tube settings.ConclusionsThe encouraging results suggest that an a priori characterization of noise for a single fluoroscopic device is feasible and could improve the actual implementation of real-time denoising strategies that take advantage of noise statistics to improve the trade-off between noise reduction and details preservation.

Influence of different wavelet bases and threshold estimation on noise reduction of substation fault signals under different SNR

With the rapid development of power industry, the number of substations is also increasing, and the voltage level is also higher and higher, and the types of equipment are diverse. The detection of equipment fault signal in substation is an important part to ensure the normal operation of electric power. Only timely detection of fault signal and corresponding processing can ensure the overall operation of substation. But the fault signal often contains a lot of white noise, which will interfere with our judgment of the fault point. How to extract the effective signal from the noisy signal becomes particularly important, but under different signal-to-noise ratio, the selection of different wavelet basis and threshold estimation also have a great impact on the noise reduction effect. In this paper, we choose different wavelet bases and threshold estimation under several different signal-to-noise ratios to study their influence on the effect of signal denoising.

Design and Analysis of a Continuous and Non-Invasive Multi-Wavelength Optical Sensor for Measurement of Dermal Water Content

Dermal water content is an important biophysical parameter in preserving skin integrity and preventing skin damage. Traditional electrical-based and open-chamber evaporimeters have several well-known limitations. In particular, such devices are costly, sizeable, and only provide arbitrary outputs. They also do not permit continuous and non-invasive monitoring of dermal water content, which can be beneficial for various consumer, clinical, and cosmetic purposes. We report here on the design and development of a digital multi-wavelength optical sensor that performs continuous and non-invasive measurement of dermal water content. In silico investigation on porcine skin was carried out using the Monte Carlo modeling strategy to evaluate the feasibility and characterize the sensor. Subsequently, an in vitro experiment was carried out to evaluate the performance of the sensor and benchmark its accuracy against a high-end, broad band spectrophotometer. Reference measurements were made against gravimetric analysis. The results demonstrate that the developed sensor can deliver accurate, continuous, and non-invasive measurement of skin hydration through measurement of dermal water content. Remarkably, the novel design of the sensor exceeded the performance of the high-end spectrophotometer due to the important denoising effects of temporal averaging. The authors believe, in addition to wellbeing and skin health monitoring, the designed sensor can particularly facilitate disease management in patients presenting diabetes mellitus, hypothyroidism, malnutrition, and atopic dermatitis.

Simultaneous identification of groundwater contamination source and aquifer parameters with a new weighted-average wavelet variable-threshold denoising method.

This paper first proposed a parallel heuristic search strategy for simultaneous identification of groundwater contamination source and aquifer parameters. As identification results are influenced by many factors, such as noisy contamination concentration data, data denoising is necessary. The existing wavelet threshold denoising method has unavoidable shortcomings; therefore, this paper first proposed a new weighted-average wavelet variable-threshold denoising (WWVD) method to improve the denoising effect for concentration data, which further enhanced the subsequent identification accuracy. However, frequent calls to the simulation model could produce high computational cost during likelihood calculation. Hence, single surrogate model of the simulation model was developed to reduce cost; however, it presented limitation. Thus, this paper first developed a differential evolution-tabu search (DE-TS) hybrid algorithm to construct an optimal ensemble surrogate model, which assembled Gaussian process, kernel extreme learning machine, and support vector regression. The first proposed DE-TS algorithm also improved the approximation accuracy of surrogate model to simulation model. This paper first proposed and implemented a parallel heuristic search iterative process for simultaneous identification, and the identification results were obtained when the iteration process terminated. The accuracy and efficiency of these newly proposed approaches were tested through a hypothetical case. Results showed that the WWVD method not only improved the denoising effect for concentration data but also enhanced the subsequent identification accuracy. The OES model using DE-TS hybrid algorithm improved the approximation accuracy of surrogate model to simulation model, and the parallel heuristic search strategy is helpful for simultaneous identification of groundwater contamination source and aquifer parameters.

Denoising convolutional neural network inspired via multi-layer convolutional sparse coding

Sparse prior to image denoising is a classical research field with a long history in computer vision. We propose an end-to-end supervised neural network, named DnMLCSC-net, which is inspired via multi-layer convolutional sparse coding model embedded with symbiotic analysis–synthesis priors for natural image denoising. Unfolding a multi-layer, learned iterative soft thresholding algorithm (ML-LISTA) and developing into a convolutional recurrent neural network, all parameters in the model are updated adaptively to minimize mixed loss via gradient descent using backpropagation. In addition, a combined ReLU function is taken as the activation function. Inconsistent dilated convolution and batch normalization were empirically introduced into the encoding layers corresponding to the first iteration of ML-LISTA. Experimental results show that our network achieves a competitive denoising effect in comparison with several state-of-the-art denoising methods.

Medical image denoising using wavelet transform and singular value decomposition

Filtering noise to recreate a high-quality image in medical image processing is an important task. During acquisition, transmission, and retrieval from storage devices, generally images are getting corrupted. So, for further analysis images must get denoised. The noises can be categorised into different types based on their nature and origin. Researchers are still looking for the effective denoising technique. Wavelet Transform (WT) is an effective transform method for denoising. Similarly Singular Value Decomposition (SVD) is also an important tool for denoising. Combining WT with SVD results in further reduction of noise. This paper proposes use of WT along with SVD for medical image denoising. Performance of image denoising is evaluated on the basis of Signal to Noise Ratio (SNR) and Peak Signal-Noise Ratio (PSNR). In the proposed approach, experimental results of WT-SVD combination gives better SNR and PSNR values than WT and SVD, if used independently.

A denoising method for pavement 3d data based on breakpoint interpolation and reference plane filtering

In order to meet the current pavement intelligent detection and feature parameter extraction requirements, a 3D pavement surface data denoising method based on breakpoint interpolation and reference plane filtering (BI-RPF) is proposed and demonstrated using 3D pavement data acquired by a laser-based 3D imaging system. First, the principal structure and major characteristics of 3D pavement data were analyzed. Next, the breakpoint interpolation method considering the relationship between ordinate and slope was proposed to complete the missing values from the original pavement data. Then, reference plane filtering based on the horizontal and longitudinal reference plane was realized so as to overcome the unevenness of the pavement. Finally, the filtering effectiveness of the proposed method was compared with that of the standard deviation filtering and median filtering methods. The results show that the proposed filtering method takes breakpoints and uneven road surface into account, which can provide more accurate 3D road surface data under more complex conditions. The algorithm provided a good denoising effect while retaining valid data describing the road surface morphology, which establishes a foundation of BI-RPF to 3D pavement reconstruction.

Analysis and comparison of Gaussian noise denoising algorithms

Collecting and processing various images has become an irreversible trend. It is meaningful to conduct more in-depth research on image denoising algorithms. We proposed mean filtering, median filtering, wiener filtering and wavelet filtering to denoise the image with Gaussian noise separately. And the objective image quality assessments are used to evaluate the quality of the images after denoising. Among them, wavelet filtering and Wiener filtering have a better effect on weaken Gaussian noise. Mean filtering and median filtering can also weaken Gaussian noise to some extent but the effect is limited. At the same time, it is equally important to select the appropriate denoising block for the diverse mean and variance of Gaussian noise. In wavelet filtering, the number of layers to be decomposed and the choice of threshold will also affect the effect of image denoising.

Estimation of Vehicle Attitude, Acceleration, and Angular Velocity Using Convolutional Neural Network and Dual Extended Kalman Filter.

The acceleration of a vehicle is important information in vehicle states. The vehicle acceleration is measured by an inertial measurement unit (IMU). However, gravity affects the IMU when there is a transition in vehicle attitude; thus, the IMU produces an incorrect signal output. Therefore, vehicle attitude information is essential for obtaining correct acceleration information. This paper proposes a convolutional neural network (CNN) for attitude estimation. Using sequential data of a vehicle’s chassis sensor signal, the roll and pitch angles of a vehicle can be estimated without using a high-cost sensor such as a global positioning system or a six-dimensional IMU. This paper also proposes a dual-extended Kalman filter (DEKF), which can accurately estimate acceleration/angular velocity based on the estimated roll/pitch information. The proposed method is validated by real-car experiment data and CarSim, a vehicle simulator. It accurately estimates the attitude estimation with limited sensors, and the exact acceleration/angular velocity is estimated considering the roll and pitch angle with de-noising effect. In addition, the DEKF can improve the modeling accuracy and can estimate the roll and pitch rates.

A cascaded convolutional neural network architecture for despeckling OCT images

Optical Coherence Tomography (OCT) is an imaging technique widely used for medical imaging. Noise in an OCT image generally degrades its quality, thereby obscuring clinical features and making the automated segmentation task suboptimal. Obtaining higher quality images requires sophisticated equipment and technology, available only in selected research settings, and is expensive to acquire. Developing effective denoising methods to improve the quality of the images acquired on systems currently in use has potential for vastly improving image quality and automated quantitative analysis. Noise characteristics in images acquired from machines of different makes and models may vary. Our experiments show that any single state-of-the-art method for noise reduction fails to perform equally well on images from various sources. Therefore, detailed analysis is required to determine the exact noise type in images acquired using different OCT machines. In this work we studied noise characteristics in the publicly available DUKE and OPTIMA datasets to build a more efficient model for noise reduction. These datasets have OCT images acquired using machines of different manufacturers. We further propose a patch-wise training methodology to build a system to effectively denoise OCT images. We have performed an extensive range of experiments to show that the proposed method performs superior to other state-of-the-art-methods.

A novel de‐noising method based on coherence average for ultrasonic signal of partial discharge in transformer

Ultrasonic method is widely used for the detection and location of partial discharge (PD) in transformer, however, the measured ultrasonic signal is usually corrupted by noise, and sometimes is even buried by noise entirely. Therefore, the de-noising of the measured signal is essential. Conventional de-nosing methods, such as wavelet method and singular value decomposition (SVD) method, generally require empirical parameter selection or estimation, and the best parameters for de-noising vary with the PD source and noise condition, which will bring some limitation to their applicability. More importantly, conventional de-noising methods usually have poor performance for the low signal-to-noise-ratio (SNR) signals. To improve this problem, the paper proposes a de-noising method based on coherence average for the ultrasonic signal. The de-noising performance of the proposed method is evaluated based on a comparison with conventional methods, the results indicate that the proposed method has great de-noising effect for the ultrasonic signals, even for the low-SNR signals. Besides, the proposed method is free from empirical parameter selection or estimation, which can make its applicability more extensive. The proposed method can offer a practical and effective solution to the de-noising of ultrasonic signal of PD in transformer.

Evaluation of denoising strategies for task-based functional connectivity: Equalizing residual motion artifacts between rest and cognitively demanding tasks.

In‐scanner head motion represents a major confounding factor in functional connectivity studies and it raises particular concerns when motion correlates with the effect of interest. One such instance regards research focused on functional connectivity modulations induced by sustained cognitively demanding tasks. Indeed, cognitive engagement is generally associated with substantially lower in‐scanner movement compared with unconstrained, or minimally constrained, conditions. Consequently, the reliability of condition‐dependent changes in functional connectivity relies on effective denoising strategies. In this study, we evaluated the ability of common denoising pipelines to minimize and balance residual motion‐related artifacts between resting‐state and task conditions. Denoising pipelines—including realignment/tissue‐based regression, PCA/ICA‐based methods (aCompCor and ICA‐AROMA, respectively), global signal regression, and censoring of motion‐contaminated volumes—were evaluated according to a set of benchmarks designed to assess either residual artifacts or network identifiability. We found a marked heterogeneity in pipeline performance, with many approaches showing a differential efficacy between rest and task conditions. The most effective approaches included aCompCor, optimized to increase the noise prediction power of the extracted confounding signals, and global signal regression, although both strategies performed poorly in mitigating the spurious distance‐dependent association between motion and connectivity. Censoring was the only approach that substantially reduced distance‐dependent artifacts, yet this came at the great cost of reduced network identifiability. The implications of these findings for best practice in denoising task‐based functional connectivity data, and more generally for resting‐state data, are discussed.

A sparse measurement matrix based method for feature enhancement of bearing fault signal

In the process of bearing signal acquisition, the overloaded data often causes great pressure on the equipment. In order to solve this problem, compressed sensing theory is used to collect signals, and the measurement matrix is sparsed into a sparse measurement matrix and a base measurement matrix. The fusion of the sparse dictionary and the measurement matrix expands the selections of the measurement matrix and improves the reconstruction accuracy. For the phenomenon of the local aggregation of matrix energy caused by the initialization of the sparse measurement matrix, the method of segmented reconstruction is proposed to solve it. Then, in view of the situation that the useful information in the reconstructed signal is polluted by a lot of noise, a new wavelet threshold denoising method is proposed, which is fused with the empirical wavelet transform (EWT), and the margin index and the peak-to-peak value are used as sensitive parameters to adjust the threshold, the effect of denoising and fault feature enhancement are performed simultaneously. Through two-step strategy, the effective acquisition of bearing signals and the enhancement of fault features are realized. The performance of this method is verified by simulation and actual bearing data and the results are encouraging.

Image denoising based on BCOLTA: Dataset and study

Robot deburring is an effective method for improving the surface quality of the high-voltage copper contact. The first step of robot deburring is to acquire the burr images. We propose a new burr mathematical model and build a real burr image dataset for burr image denoising. In order to improve burr image denoising effects of the high-voltage copper contact, this study proposes an online burr image denoising algorithm, that is, block cosparsity overcomplete learning transform algorithm (BCOLTA). The penalty term and the condition number are affected by the burr parameter. The clustering and transform alternate minimisation algorithms are adopted to achieve lower computational cost and better denoising effect. In addition, BCOLTA also has a good adaptibility to inherent noise images, especially in Gaussian noise. Compared with other traditional and deep learning algorithms by no reference and full reference image quality assessment methods, BCOLTA has state-of-the-art denoising effects and computational complexity on dealing with burr images. This research will play an important role in the intelligent manufacturing field.

Image processing effects on the deep face recognition system.

Face recognition technology has become an important quantitative examination method in the field of forensic identification of human images. However, face image quality affects the recognition performance of face recognition systems. Existing research on the effects of face image denoising and enhancement methods on the face recognition performance are typically based on facial images with manually synthesized noises rather than the noises under natural environmental corruption, and their studied face recognition techniques are limited on the traditional face recognition algorithms rather than state-of-the-art convolutional neural network based face recognition methods. In this work, face image materials from 33 real cases in forensic identification of human images were collected for quantitative analysis of the effects of face image denoising and enhancement methods on the deep face recognition performance of the MXNet system architecture based face recognition system. The results show that face image quality has a significant effect on the recognition performance of the face recognition system, and the image processing techniques can enhance the quality of face images, and then improve the recognition precision of the face recognition system. In addition, the effects of the Gaussian filtering are better than the self-snake model based image enhancement method, which indicates that the image denoising methods are more suitable for performance improvement of the deep face recognition system rather than the image enhancement techniques under the application of the practical cases.