Remove Noise Components Research Articles

Multisensor signal denoising based on matching synchrosqueezing wavelet transform for mechanical fault condition assessment

Since it is difficult to obtain the accurate running status of mechanical equipment with only one sensor, multisensor measurement technology has attracted extensive attention. In the field of mechanical fault diagnosis and condition assessment based on vibration signal analysis, multisensor signal denoising has emerged as an important tool to improve the reliability of the measurement result. A reassignment technique termed the synchrosqueezing wavelet transform (SWT) has obvious superiority in slow time-varying signal representation and denoising for fault diagnosis applications. The SWT uses the time–frequency reassignment scheme, which can provide signal properties in 2D domains (time and frequency). However, when the measured signal contains strong noise components and fast varying instantaneous frequency, the performance of SWT-based analysis still depends on the accuracy of instantaneous frequency estimation. In this paper, a matching synchrosqueezing wavelet transform (MSWT) is investigated as a potential candidate to replace the conventional synchrosqueezing transform for the applications of denoising and fault feature extraction. The improved technology utilizes the comprehensive instantaneous frequency estimation by chirp rate estimation to achieve a highly concentrated time–frequency representation so that the signal resolution can be significantly improved. To exploit inter-channel dependencies, the multisensor denoising strategy is performed by using a modulated multivariate oscillation model to partition the time–frequency domain; then, the common characteristics of the multivariate data can be effectively identified. Furthermore, a modified universal threshold is utilized to remove noise components, while the signal components of interest can be retained. Thus, a novel MSWT-based multisensor signal denoising algorithm is proposed in this paper. The validity of this method is verified by numerical simulation, and experiments including a rolling bearing system and a gear system. The results show that the proposed multisensor matching synchronous squeezing wavelet transform (MMSWT) is superior to existing methods.

Voiced/unvoiced speech classification‐based adaptive filtering of decomposed empirical modes for speech enhancement

This study presents a speech filtering method exploiting the combined effects of the empirical mode decomposition (EMD) and the local statistics of the speech signal using the adaptive centre weighted average (ACWA) filter. The novelty lies in incorporating the frame class (voiced/unvoiced) in the conventional filtering using the EMD and the ACWA filter. The speech signal is segmented into frames and each one is broken down by the EMD into a finite number of intrinsic mode functions (IMFs). The number of filtered IMFs depends on whether the frame is voiced or unvoiced. An energy criterion is used to identify voiced frames while a stationarity index distinguishes between unvoiced and transient sequences. Reported results obtained on signals corrupted by additive noise (white, F16, factory) show that the proposed filtering in line with the frame class is very effective in removing noise components from noisy speech signal. Compared with filtering results of the wavelet, the ACWA, and the EMD-ACWA methods, the proposed technique gives much better results in terms of average segmental signal-to-noise ratio and listening quality based on perceptual evaluation speech quality score.

De-noising of ECG Signal and QRS Detection Using Hilbert Transform and Adaptive Thresholding

Electrocardiogram (ECG) is a primary diagnostic tool for cardiac disorders. During acquisition of ECG signals different noises like instrument noise, muscle noise, motion artifacts and baseline wander are frequently mixed with signals in real-time situation. The segmentation and detection of R peaks in the ECG is the initial steps in HRV analysis. In this paper, we employ the discrete wavelet transform to remove noise components of the time - frequency domain in order to enhance the ECG signal and the Hilbert transform with the adaptive thresholding technique used to explore an optimal combination to detect R-peaks more accurately. The proposed method is evaluated on ECG signals from MIT database. The experimental results of present method show better signal to noise ratio (SNR) with lower mean square error (MSE). To evaluate the quality of physiological information preserved in the enhanced ECG signal, the R-peak detection was also tested. The performance of the proposed method is found to be better in detecting R-peaks having sensitivity of 99.71% and the positive predictability of 99.72% respectively with less detection error rate of 0.52%.

ICA-based artifact removal diminishes scan site differences in multi-center resting-state fMRI.

Resting-state fMRI (R-fMRI) has shown considerable promise in providing potential biomarkers for diagnosis, prognosis and drug response across a range of diseases. Incorporating R-fMRI into multi-center studies is becoming increasingly popular, imposing technical challenges on data acquisition and analysis, as fMRI data is particularly sensitive to structured noise resulting from hardware, software, and environmental differences. Here, we investigated whether a novel clean up tool for structured noise was capable of reducing center-related R-fMRI differences between healthy subjects. We analyzed three Tesla R-fMRI data from 72 subjects, half of whom were scanned with eyes closed in a Philips Achieva system in The Netherlands, and half of whom were scanned with eyes open in a Siemens Trio system in the UK. After pre-statistical processing and individual Independent Component Analysis (ICA), FMRIB's ICA-based X-noiseifier (FIX) was used to remove noise components from the data. GICA and dual regression were run and non-parametric statistics were used to compare spatial maps between groups before and after applying FIX. Large significant differences were found in all resting-state networks between study sites before using FIX, most of which were reduced to non-significant after applying FIX. The between-center difference in the medial/primary visual network, presumably reflecting a between-center difference in protocol, remained statistically significant. FIX helps facilitate multi-center R-fMRI research by diminishing structured noise from R-fMRI data. In doing so, it improves combination of existing data from different centers in new settings and comparison of rare diseases and risk genes for which adequate sample size remains a challenge.

Generalized maximum a posteriori spectral amplitude estimation for speech enhancement

Spectral restoration methods for speech enhancement aim to remove noise components in noisy speech signals by using a gain function in the spectral domain. How to design the gain function is one of the most important parts for obtaining enhanced speech with good quality. In most studies, the gain function is designed by optimizing a criterion based on some assumptions of the noise and speech distributions, such as minimum mean square error (MMSE), maximum likelihood (ML), and maximum a posteriori (MAP) criteria. The MAP criterion shows advantage in obtaining a more reliable gain function by incorporating a suitable prior density. However, it has a problem as several studies showed: although MAP based estimator effectively reduces noise components when the signal-to-noise ratio (SNR) is low, it brings large speech distortion when the SNR is high. For solving this problem, we have proposed a generalized maximum a posteriori spectral amplitude (GMAPA) algorithm in designing a gain function for speech enhancement. The proposed GMAPA algorithm dynamically specifies the weight of prior density of speech spectra according to the SNR of the testing speech signals to calculate the optimal gain function. When the SNR is high, GMAPA adopts a small weight to prevent overcompensations that may result in speech distortions. On the other hand, when the SNR is low, GMAPA uses a large weight to avoid disturbance of the restoration caused by measurement noises. In our previous study, it has been proven that the weight of the prior density plays a crucial role to the GMAPA performance, and the weight is determined based on the SNR in an utterance-level. In this paper, we propose to compute the weight with the consideration of time–frequency correlations that result in a more accurate estimation of the gain function. Experiments were carried out to evaluate the proposed algorithm on both objective tests and subjective tests. The experimental results obtained from objective tests indicate that GMAPA is promising compared to several well-known algorithms at both high and low SNRs. The results of subjective listening tests indicate that GMAPA provides significantly higher sound quality than other speech enhancement algorithms.

Sparse Representation-Based ISAR Imaging Using Markov Random Fields

To encourage the continuity of the target scene, a novel sparse representation (SR)-based inverse synthetic aperture radar (ISAR) imaging algorithm is proposed by leveraging the Markov random fields (MRF). The ISAR imaging problem is reformulated in a Bayesian framework where correlated priors are used for the hidden variables to enforce the continuity of target scene. To further enforce the nonzero or zero scatterers to cluster in a spatial consistent manner, the MRF is used as the prior for the support of the target scene. To surmount the difficulty of calculating the posterior due to the imposed correlated priors and the MRF, variational Bayes expectation-maximization (VBEM) method is used to simultaneously approximate the posterior of the hidden variables and estimate the model parameters of the MRF. The convergence of the method is easily diagnosed by commonly used stopping criterion. Both the synthetic and the experimental results demonstrate that the proposed algorithm can achieve substantial improvements in terms of preserving the weak scatterers and removing noise components over other reported SR-based ISAR imaging algorithms.

Pathway-based factor analysis of gene expression data produces highly heritable phenotypes that associate with age.

Statistical factor analysis methods have previously been used to remove noise components from high-dimensional data prior to genetic association mapping and, in a guided fashion, to summarize biologically relevant sources of variation. Here, we show how the derived factors summarizing pathway expression can be used to analyze the relationships between expression, heritability, and aging. We used skin gene expression data from 647 twins from the MuTHER Consortium and applied factor analysis to concisely summarize patterns of gene expression to remove broad confounding influences and to produce concise pathway-level phenotypes. We derived 930 “pathway phenotypes” that summarized patterns of variation across 186 KEGG pathways (five phenotypes per pathway). We identified 69 significant associations of age with phenotype from 57 distinct KEGG pathways at a stringent Bonferroni threshold (). These phenotypes are more heritable () than gene expression levels. On average, expression levels of 16% of genes within these pathways are associated with age. Several significant pathways relate to metabolizing sugars and fatty acids; others relate to insulin signaling. We have demonstrated that factor analysis methods combined with biological knowledge can produce more reliable phenotypes with less stochastic noise than the individual gene expression levels, which increases our power to discover biologically relevant associations. These phenotypes could also be applied to discover associations with other environmental factors.

Research on Blasting Vibration Signal Denoising Based on Lifting Scheme

Considering lifting scheme and traditional wavelet packet transform principle, The optimal lifting wavelet packet threshold denoising algorithm was introduced. Experimental blasting vibration signal was decomposed by optimal lifting wavelet packet, and noise components in blasting vibration measured signals were filtered successfully. Research shows that, lifting wavelet package transform can effectively remove noise components, and it laid an important foundation for lifting algorithm will be introduced into the analysis field of blasting vibration effects and other mechanical vibration signal.

A Class of Multivariate Denoising Algorithms Based on Synchrosqueezing

Univariate thresholding techniques based on high resolution time-frequency algorithms, such as the synchrosqueezing transform, have emerged as important tools in removing noise from real world data. Low cost multichannel sensor technology has highlighted the need for direct multivariate denoising, and to this end, we introduce a class of multivariate denoising techniques based on the synchrosqueezing transform. This is achieved by partitioning the time-frequency domain so as to identify a set of modulated oscillations common to the constituent data channels within multivariate data, and by employing a modified universal threshold in order to remove noise components, while retaining signal components of interest. This principle is used to introduce both the wavelet and Fourier based multivariate synchrosqueezing denoising algorithms. The performance of the proposed multivariate denoising algorithm is illustrated on both synthetic and real world data.

PROJECTED PRINCIPAL COMPONENT ANALYSIS IN FACTOR MODELS.

This paper introduces a Projected Principal Component Analysis (Projected-PCA), which employees principal component analysis to the projected (smoothed) data matrix onto a given linear space spanned by covariates. When it applies to high-dimensional factor analysis, the projection removes noise components. We show that the unobserved latent factors can be more accurately estimated than the conventional PCA if the projection is genuine, or more precisely, when the factor loading matrices are related to the projected linear space. When the dimensionality is large, the factors can be estimated accurately even when the sample size is finite. We propose a flexible semi-parametric factor model, which decomposes the factor loading matrix into the component that can be explained by subject-specific covariates and the orthogonal residual component. The covariates' effects on the factor loadings are further modeled by the additive model via sieve approximations. By using the newly proposed Projected-PCA, the rates of convergence of the smooth factor loading matrices are obtained, which are much faster than those of the conventional factor analysis. The convergence is achieved even when the sample size is finite and is particularly appealing in the high-dimension-low-sample-size situation. This leads us to developing nonparametric tests on whether observed covariates have explaining powers on the loadings and whether they fully explain the loadings. The proposed method is illustrated by both simulated data and the returns of the components of the S&P 500 index.

ECG signal enhancement using S-Transform

Electrocardiogram (ECG), which is a noninvasive technique, is used generally as a primary diagnostic tool for cardiovascular diseases. In real-time scenario, noises like channel noise, muscle artifacts, electrode motion and baseline wander are often embedded with ECG signals during acquisition and transmission. In this paper, an automatic ECG signal enhancement technique is proposed to remove noise components from time–frequency domain represented noisy ECG signal. Stockwell transform (S-Transform) is used in this work to represent the noisy ECG signal in time–frequency domain. Next, masking and filtering technique is applied to remove unwanted noise components from time–frequency domain. The proposed technique does not require any prior information like R-peak position or reference signal as auxiliary signal. This method is evaluated on ECG signals which are available in MIT-BIH Arrhythmia database. The experimental results demonstrate that the proposed method shows better signal to noise ratio (SNR) and lower root means square error (RMSE) compared to earlier reported wavelet transform with soft thresholding (WT-Soft) and wavelet transform with subband dependent threshold (WT-Subband) based technique. To quantify the significant difference among all methods, the performances of different ECG enhancement techniques at 1.25dB input SNR level are compared using analysis of variance (ANOVA) based statistical evaluation technique and it is seen that the proposed method yields superior performance compared to other methods. R-peak detection test is also conducted on enhanced ECG signal in addition to SNR and RMSE to evaluate the quality of biology-related information preserved in the enhanced ECG signal. The performance of R-peak detection for denoised ECG signals, in terms of sensitivity and positive predictivity using proposed enhancement method, is also better than WT-Soft, WT-Subband methods, and validates the superiority of the proposed method.

유도전동기를 위한 고 신뢰성 고장 검출 및 분류 알고리즘 연구

This paper proposes a 3-stage (preprocessing, feature extraction, and classification) fault detection and classification algorithm for induction motors. In the first stage, a low-pass filter is used to remove noise components in the fault signal. In the second stage, a discrete cosine transform (DCT) and a statistical method are used to extract features of the fault signal. Finally, a back propagation neural network (BPNN) method is applied to classify the fault signal. To evaluate the performance of the proposed algorithm, we used one second long normal/abnormal vibration signals of an induction motor sampled at 8kHz. Experimental results showed that the proposed algorithm achieves about 100% accuracy in fault classification, and it provides 50% improved accuracy when compared to the existing fault detection algorithm using a cross-covariance method. In a real-world data acquisition environment, unnecessary noise components are usually included to the real signal. Thus, we conducted an additional simulation to evaluate how well the proposed algorithm classifies the fault signals in a circumstance where a white Gaussian noise is inserted into the fault signals. The simulation results showed that the proposed algorithm achieves over 98% accuracy in fault classification. Moreover, we developed a testbed system including a TI`s DSP (digital signal processor) to implement and verify the functionality of the proposed algorithm.

Combining Speech Enhancement and Cepstral Mean Normalization for LPC Cepstral Coefficients

A mismatch between the training and testing in noisy circumstance often causes a drastic decrease in the performance of speech recognition system. The robust feature coefficients might suppress this sensitivity of mismatch during the recognition stage. In this paper, we investigate the noise robustness of LPC Cepstral Coefficients (LPCC) by using speech enhancement with feature post-processing. At front-end, speech enhancement in the wavelet domain is used to remove noise components from noisy signals. This enhanced processing adopts the combination of discrete wavelet transform (DWT), wavelet packet decomposition (WPD), multi-thresholds processing etc to obtain the estimated speech. The feature post-processing employs cepstral mean normalization (CMN) to compensate the signal distortion and residual noise of enhanced signals in the cepstral domain. The performance of digit speech recognition systems is evaluated under noisy environments based on NOISEX-92 database. The experimental results show that the presented method exhibits performance improvements in the adverse noise environment compared with the previous features.

다양한 실내 환경변수로부터 강인한 객체 검출

본 논문에서는 다양한 환경변수가 존재하는 실내에서 효율적인 객체를 검출하기 위한 방법을 제안한다. 일반적으로 실내 환경은 조명의 변화와 객체에 의해 발생된 그림자, 바닥면에 반사된 조명성분 등으로 인하여 정확한 객체 검출이 이루어지기 어려운 환경이다. 먼저 객체검출을 위한 배경영상을 생성한다. 영상 내에 객체가 존재하는 경우 이전에 생성된 배경영상과 현재 입력영상간의 유사도 비교를 통해 보정된 배경영상을 생성한다. 배경영상과 입력영상으로 생성한 평균영상과 보정된 배경영상을 이용하여 혼합영상을 생성한다. 마지막으로 혼합영상을 이용하여 입력된 영상으로부터 객체를 검출한다. 검출된 객체를 보완하기 위해 레이블링 과정을 통해 잡음 성분을 제거한 후 모폴로지 기법을 적용하여 객체영역 보완한다. 따라서 조명의 변화나 그림자와 같은 환경변수로부터 강인한 객체를 검출한다. 본 논문에서 제안한 시스템은 변형된 조명성분과 그림자 성분이 포함되어 있는 혼합영상을 사용하기 때문에 기존시스템보다 객체영역 검출이 더욱 효과적이다. In this paper, we propose a detection method of reduced computational complexity aimed at separating the moving objects from the background in a generic video sequence. In generally, indoor environments, it is difficult to accurately detect the object because environmental factors, such as lighting changes, shadows, reflections on the floor. First, the background image to detect an object is created. If an object exists in video, on a previously created background images for similarity comparison between the current input image and to detect objects through several operations to generate a mixture image. Mixed-use video and video inputs to detect objects. To complement the objects detected through the labeling process to remove noise components and then apply the technique of morphology complements the object area. Environment variable such as, lighting changes and shadows, to the strength of the object is detected. In this paper, we proposed that environmental factors, such as lighting changes, shadows, reflections on the floor, including the system uses mixture images. Therefore, the existing system more effectively than the object region is detected.

Development of Gradient Descent Adaptive Algorithms to Remove Common Mode Artifact for Improvement of Cardiovascular Signal Quality

Common-mode noise degrades cardiovascular signal quality and diminishes measurement accuracy. Filtering to remove noise components in the frequency domain often distorts the signal. Two adaptive noise canceling (ANC) algorithms were tested to adjust weighted reference signals for optimal subtraction from a primary signal. Update of weight w was based upon the gradient term of the steepest descent equation: [see text], where the error epsilon is the difference between primary and weighted reference signals. nabla was estimated from Deltaepsilon(2) and Deltaw without using a variable Deltaw in the denominator which can cause instability. The Parallel Comparison (PC) algorithm computed Deltaepsilon(2) using fixed finite differences +/- Deltaw in parallel during each discrete time k. The ALOPEX algorithm computed Deltaepsilon(2)x Deltaw from time k to k + 1 to estimate nabla, with a random number added to account for Deltaepsilon(2) . Deltaw--> 0 near the optimal weighting. Using simulated data, both algorithms stably converged to the optimal weighting within 50-2000 discrete sample points k even with a SNR = 1:8 and weights which were initialized far from the optimal. Using a sharply pulsatile cardiac electrogram signal with added noise so that the SNR = 1:5, both algorithms exhibited stable convergence within 100 ms (100 sample points). Fourier spectral analysis revealed minimal distortion when comparing the signal without added noise to the ANC restored signal. ANC algorithms based upon difference calculations can rapidly and stably converge to the optimal weighting in simulated and real cardiovascular data. Signal quality is restored with minimal distortion, increasing the accuracy of biophysical measurement.

Adaptive filtering using harmonic structure of voiced speech for reducing nonstationary known noise

Proposed is an effective method for reducing nonstationary known noise. The objective of this research is to develop a scheme of preprocessing for speech recognition that keeps the same speech recognition rate even in a worse acoustic environment and to realize a TV control system using speech recognition. The basic idea of the proposed method is to estimate the frequency spectrum of noise including sounds from the TV itself and to remove noise components from the frequency spectrum of the received signal. A transfer function from the TV set to the microphone is calculated in an iterative manner estimating the noise signal at the microphone. Traditional ANC techniques do not directly use spectral features such as harmonic structure or fundamental frequency, for example. As the proposed method uses harmonic structure of vocalic segments in command speech, it is expected to have an advantage compared with the traditional adaptive methods. Results of evaluation using speech recognition show that the proposed method significantly improves speech recognition rate compared with conventional spectral subtraction. [Work supported by Knowledge Cluster Project, MEXT, and by Academic Frontier Project Doshisha University.]

三次元補償コイルとＩＣＡを用いた生体磁気計測における雑音低減の検討

In biomagnetic measurement, biomagnetic signals are extremely weak in comparison with environmental magnetic noise. There are several hardware and software noise-reduction methods, such as the use of a gradiometer, averaging,filtering, the PCA method, and independent component analysis(ICA). The ICA method is used to extract and remove noise components from the brain’s magnetic field and the signal. We carried out a 3-D MEG vector measurement of a somatosensory evoked field(SEF) using a 39-channel SQUID system. To improve the signal-to-noise ratio of SEF data, we propose a new method, combining a reference coil and ICA, and apply it to SEF data.

A Noise Reduction Method Based on Linear Prediction with Variable Step-Size

A noise reduction technique that uses the linear prediction to remove noise components in speech signals has been proposed previously. The noise reduction works well for additive white noise signals, because the coefficients of the linear predictor converge such that the prediction error becomes white. In this method, the linear predictor is updated by a gradient-based algorithm with a fixed step-size. However, the optimal value of the step-size changes with the values of the prediction coefficients. In this paper, we propose a noise reduction system using the linear predictor with a variable step-size. The optimal value of the step-size depends also on the variance of the white noise, however the variance is unknown. We therefore introduce a speech/non-speech detector, and estimate the variance in non-speech segments where the observed signal includes only noise components. The simulation results show that the noise reduction capability of the proposed system is better than that of the conventional one with a fixed step-size.

A study on noise reduction using ICA for Magnetoencephalography

In the biomagnetic measurement, the biomagnetic signal is extremely weak compared with environmental magnetic noise. Therefore, it is important to reduce the noise component. There are many noise-reduction studies for MEG using Independent Component Analysis (ICA). The ICA method is expectable to extract and remove noise components from the brain magnetic field measurement data. However, in these researches, each obtained independent components are artificially distinguished to the noise and the signal. We propose a method of distinguishing to the noise and the signal automatically by using the signal subspace method for vector brain magnetic field. By applying this method to a phantom data and Auditory Evoked Field data, it is shown improvement of the signal to noise ratio and estimated accuracy.

Enhancement of electrolarynx speech using adaptive noise cancelling based on independent component analysis.

The electrolarynx provides a valuable means of verbal communication for people who cannot use their natural voice-production mechanism, but technology has changed very little since it was introduced in the 1950s. The presence of background noise degrades the resulting speech. In this study background noise was reduced by a new method, independent component analysis-based adaptive noise cancelling, which can remove noise components of the primary input signal based on statistical independence, by incorporating both second-order and higher-order statistics. The method shows better performance than the conventional least mean square algorithm. Acoustic analysis of the denoised electrolarynx speech revealed a significant reduction in the amount of background noise. Results from the perceptual evaluations indicated that the new filtering technique produced a noticeable improvement in the acceptability of the electrolarynx speech in a quiet environment (from 1.75 to 2.49, arbitrary units) or a noisy environment (from 0.59 to 1.82). In general, there was no significant improvement or degradation in intelligibility in the quiet environment (from 52.7 to 53.3). However, the processing did improve the intelligibility in a babble-noise environment (from 24.9 to 40.6). The improvement in acceptability and intelligibility may increase the communication ability of the user in daily situations.