基于中值滤波和提升小波分析的图像去噪方法研究

针对现有算法大多对单一高斯噪声或脉冲噪声进行图像滤波的问题,在对二维图像平滑去噪的过程中,采用基于中值滤波和提升小波变换相结合的图像去噪方法。在中值滤波基础上,构造基于脉冲检测的中值滤波器,找出混合噪声中脉冲噪声并进行滤波;与此同时,对原始小波进行提升,构造提升小波,然后采用提升小波阈值去噪方法抑制高斯噪声。实验结果表明:采用本文方法,混合噪声得到有效抑制,去噪效果好。     

Preferred listening level for speech disturbed by fluctuating noise

This paper is a sequel to an earlier one in which we investigated, among other things, the influence of stationary background noise on preferred listening level. The influence of fluctuations in noise level on preferred listening level was compared with the influence of stationary noise. It turned out that the preferred listening level for a read-aloud text, adjusted by listeners against a background of noise, is hardly influenced by fluctuations in the noise, provided the equivalent noise level (L_eq) in dB(A) remains the same. This holds both for systematically modulated noise down to 0·1 Hz and more randomly fluctuating noise such as traffic noise. Average preferred listening level proved to be an accurate measure for evaluating various conditions, such as modulation frequency and noise level, in a single experiment.     

External non-white noise: Theory and experiment

We consider a system under the influence of an external gaussian noise with a flat spectrum and finite cut-off frequency. The stationery properties of the system are studied as a function of the correlation time of the noise. Theoretical predictions based on an expansion in powers of the correlation time are corroborated by experimental measurements on an electric circuit with a digital noise generator. The two main qualitative properties experimentally observed which can not be explained in a white noise limit theory are: (i) appearance of relative extrema of the stationery distribution for high noise intensity when the correlation time is not too small; (ii) dependence on the noise characteristics of the position of the maximum of the stationary distribution for low noise intensity.     

基于改进中值滤波和提升小波变换的阈值去噪方法研究

对于实际拍摄的一些图像信噪比低,噪声密度大,且含有混合噪声,而现有算法大多只能去除单一噪声的问题。针对混合噪声中含有的脉冲噪声和高斯噪声,提出基于改进中值滤波和提升小波变换去噪相结合的方法。去噪过程中,使用中值滤波器提取脉冲噪声并采用中值滤波算法滤波后,构造提升小波,采用改进阈值函数提升小波阈值去噪方法去除高斯噪声。实验结果表明,当噪声值(,)=(0.4, 20)时,采用本文去噪方法,峰值信噪比（PSNR）为34.002 1,平均绝对误差（MAE）为2.365 3。     

Effects of noise on periodic orbits of the logistic map

Noise can induce an inverse period-doubling transition and chaos. The effects of noise on each periodic orbit of three different period sequences are investigated for the logistic map. It is found that the dynamical behavior of each orbit, induced by an uncorrelated Gaussian white noise, is different in the mergence transition. For an orbit of the period-six sequence, the maximum of the probability density in the presence of noise is greater than that in the absence of noise. It is also found that, under the same intensity of noise, the effects of uncorrelated Gaussian white noise and exponentially correlated colored (Gaussian) noise on the period-four sequence are different.      

圆锯机空气动力噪声

本文通过对圆锯空载时空气动力噪声的发声机理研究，探讨这类噪声（气动噪声）的主源及其辐射特征，最后提出低噪据片的改进型结构。     

The power spectrum and correlation of flow noise for an axisymmetric body in water

Understanding the physical features of the flow noise for an axisymmetric body is important for improving the performance of a sonar mounted on an underwater platform. Analytical calculation and numerical analysis of the physical features of the flow noise for an axisymmetric body are presented and a simulation scheme for the noise correlation on the hydrophones is given. It is shown that the numerical values of the flow noise coincide well with the analytical values. The main physical features of flow noise are obtained. The flow noises of two different models are compared and a model with a rather optimal fore-body shape is given. The flow noise in horizontal symmetry profile of the axisymmetric body is non-uniform, but it is omni-directional and has little difference in the cross section of the body. The loss of noise diffraction has a great effect on the flow noise from boundary layer transition. Meanwhile, based on the simulation, the noise power level increases with velocity to approximately the fifth power at high frequencies, which is consistent with the experiment data reported in the literature. Furthermore, the flow noise received by the acoustic array has lower correlation at a designed central frequency, which is important for sonar system design.     

The effect of narrow-band noise maskers on increment detection

It is often assumed that listeners detect an increment in the intensity of a pure tone by detecting an increase in the energy falling within the critical band centered on the signal frequency. A noise masker can be used to limit the use of signal energy falling outside of the critical band, but facets of the noise may impact increment detection beyond this intended purpose. The current study evaluated the impact of envelope fluctuation in a noise masker on thresholds for detection of an increment. Thresholds were obtained for detection of an increment in the intensity of a 0.25- or 4-kHz pedestal in quiet and in the presence of noise of varying bandwidth. Results indicate that thresholds for detection of an increment in the intensity of a pure tone increase with increasing bandwidth for an on-frequency noise masker, but are unchanged by an off-frequency noise masker. Neither a model that includes a modulation-filter-bank analysis of envelope modulation nor a model based on discrimination of spectral patterns can account for all aspects of the observed data.     

Stochastic resonance in linear system driven by multiplicative and additive noise

The stochastic resonance (SR) is studied in an overdamped linear system driven by multiplicative and additive noise when the additive noise is a linear combination of an asymmetric dichotomous noise and its square. The exact expressions are obtained for the first two moments and the correlation function and the SR phenomenon appeared. There are three different forms of SR: the bona fide SR, the conventional SR and SR in the broad sense. Moreover, the asymmetry of multiplicative noise has different effect on signal-to-ratio (SNR) for the first two different forms of SR and the effects of multiplicative noise and additive noise on SNR are different.     

750 kV变电站电晕噪声的影响与防治

电晕噪声不同于的传统的电气设备噪声,其发声位置高、频带宽、声源分散,是750 kV变电站噪声控制的薄弱环节.该文以西北某750 kV变电站为例,使用紫外成像仪定位了36处主要电晕噪声点,基于噪声实测和全站声场仿真分析,研究了电晕噪声对变电站声环境的影响.针对主要电晕点的实际情况,提出了具体的电晕噪声防治措施.治理措施在...     

Entanglement dynamics in atom-field interaction in the presence of noise: a special application of the Burshtein equation

The entanglement dynamics in a system of the interaction of an atom with a single-mode thermal field in the presence of noise is studied by the Jaynes-Cummings model. Two-state random phase telegraph noise is considered as the noise in the interaction and an exact solution to the model under this noise is obtained by the Burshtein equation. Although the Burshtein equation is applicable for laser-atom interactions, it is shown that it can be applied to atom-thermal field system as a special case. The solution is used to investigate the entanglement dynamics of the atom-field interaction by calculating a lower bound on concurrence. It is found that the entanglement is a non monotonic function of the intensity of the noise. The degree of the entanglement decreases to a minimum value for an optimal intensity of the noise and then increases for a sufficiently large intensity. Moreover, intense noise may generate stronger entanglement compared with the absence of noise.     

Effect of phase noise on parametric instabilities

We report an experimental study on the effect of an external phase noise on the parametric amplification of surface waves. We observe that both the instability growth rate and the wave amplitude above the instability onset are decreased in the presence of noise. We show that all the results can be understood with a deterministic amplitude equation for the wave in which the effect of noise is just to change the forcing term. All the data for the growth rate (respectively the wave amplitude), obtained for different forcing amplitudes and different intensities of the noise, can be collapsed on a single curve using this renormalized forcing in the presence of noise.     

On the role of aerodynamically generated sound in determining wayside noise levels from high speed trains

The relative contributions of aerodynamic and wheel/rail noise to railway wayside noise levels are not well understood. Methods for predicting these contributions discussed in this paper include (i) an equation for turbulent boundary layer noise (the minimum wayside noise), (ii) an empirical formula for total aerodynamic noise based on airframe noise studies, and (iii) the Peters equation for wheel/rail interaction noise. Comparisons are made between predicted and measured noise levels for (i) a buoyant vehicle, (ii) the Linear Induction Motor Research Vehicle (LIMRV), and (iii) a magnetically levitated vehicle. Analysis of these results indicates that aerodynamic fluctuations could become the dominant source 3f wayside noise at train speeds of 240–280 km/h. This prognosis is for new high speed railway vehicles equipped with disc brakes and other innovations that reduce the wheel/rail noise contribution.     

Total annoyance from an industrial noise source with a main spectral component combined with a background noise

When living close to an industrial plant, people are exposed to a combination of industrial noise sources and a background noise composed of all the other noise sources in the environment. As a first step, noise annoyance indicators in laboratory conditions are proposed for a single exposure to an industrial noise source. The second step detailed in this paper involves determining total annoyance indicators in laboratory conditions for ambient noises composed of an industrial noise source and a background noise. Two types of steady and permanent industrial noise sources are studied: low frequency noises with a main spectral component at 100 Hz, and noises with a main spectral component in middle frequencies. Five background noises are assessed so as to take into account different sound environments which can usually be heard by people living around an industrial plant. One main conclusion of this study is that two different analyses are necessary to determine total annoyance indicators for this type of ambient noise, depending on the industrial noise source composing it. Therefore, two total annoyance indicators adapted to the ambient noises studied are proposed.     

Train noise reduction scenarios for compliance with future noise legislation

The Technical Specification for Interoperability (TSI) for high-speed trains on the European market includes limits on noise emission. These and other future restrictions on exterior noise of high-speed and intercity trains will require that train manufacturers implement noise control measures early in the design phase.A fundamental problem faced by manufacturers during the design process is determining how much noise reduction is required for each of the various noise sources on the train in order to achieve an optimal balance. To illustrate this process, estimates are presented of the contributions from different sources on existing Bombardier trains, based on measured data, numerical calculations and empirical formulae.In addition, methods of achieving the required noise reductions for different sources are briefly discussed along with targets for future exterior noise emission.Measurement results presented demonstrate the importance of track quality in noise emission. Noise restrictions, including future legislation, must give proper recognition to this important parameter.     

Phase reduction of stochastic limit cycle oscillators

We point out that for an oscillator subjected to noise the conventional phase equation is not a proper approximation even for weak noise. We present a phase reduction method valid for an oscillator subjected to weak white Gaussian noise. Numerical evidence demonstrates that the phase equation properly approximates dynamics of the original oscillator. Moreover, we show that, in general, noise causes a shift of the oscillator frequency and discuss its effects on entrainment.     

Comparative Analysis of Different Statistics for Detecting Spatial Signals in the Case of Short Samples

In this paper, we perform comparative analysis of several algorithms for signal processing in an antenna array used for determining multidimensional Gaussian complex signals with an a priori unknown spatial covariance matrix against the background of Gaussian noise. All the studied decision statistics are obtained on the basis of generalized likelihood ratio for arbitrary-size samples. Depending on the available a priori information on the received signal and noise, we consider three cases of noise background, namely, nonuniform noise over the antenna elements with unknown power, uniform noise with unknown power, and uniform noise with known (unit) power. To determine the threshold values of the decision statistics, their probability densities are expanded in a series in terms of orthogonal Jacobi polynomials. Detection curves are constructed for all considered cases on the basis of obtained thresholds. The obtained results are valid for short samples whose size is comparable with the number of elements in the receiving antenna.     

Noise sensitivity and reactions to noise and other environmental conditions

This article integrates findings from the literature and new results regarding noise sensitivity. The new results are based on analyses of 28 combined datasets (N = 23,038), and separate analyses of a large aircraft noise study (N = 10,939). Three topics regarding noise sensitivity are discussed, namely, its relationship with noise exposure, its working mechanism, and the scope of its influence. (1) A previous review found that noise sensitivity has no relationship with noise exposure. The current analyses give consistent results, and show that there is at most a very weak positive relationship. (2) It was observed earlier that noise sensitivity alters the effect of noise exposure on noise annoyance, and does not (only) have an additive effect. The current analyses confirm this, and show that the relation of the annoyance score with the noise exposure is relatively flat for nonsensitives while it is steeper for sensitives. (3) Previous studies showed that noise sensitivity also influences reactions other than noise annoyance. The current analyses of the aircraft noise study extend these results, but also indicate that noise sensitivity has relatively little influence on reactions to nonenvironmental conditions.     

Optimum splitting ratio for amplifier noise reduction by an asymmetric nonlinear optical loop mirror

We theoretically and experimentally analyze the influence of the splitting ratio and the input power on the noise reduction capability of an asymmetric nonlinear optical loop mirror (NOLM) for different input noise levels. An easy method to calculate the optimum parameters for noise reduction is also presented. The best noise reduction is found at NOLM input powers at which the nonlinear transfer function has a slope close to zero. Additionally, the splitting ratio of the NOLM has to be adapted to its input noise level to suppress amplitude fluctuations effectively. Since the noise reduction by the NOLM is due to the Kerr nonlinearity, which has a timescale below a few femtoseconds, the noise reduction is applicable to short pulses in the picosecond and femtosecond range. This makes the NOLM applicable as an optical regenerator in an optical data transmission system at high bit rates, such as 160 GBit/s.     

基于小波变换的图像混合噪声自适应滤除算法

为同时滤除图像中的椒盐噪声和高斯噪声,提出了一种基于小波变换的混合噪声自适应滤除算法,该算法首先采用中值滤波去除椒盐噪声,然后借助边缘检测算子区将图像为分边缘与非边缘区域,进一步对非边缘区域引入改进的均值滤波器,有效削弱高斯噪声的同时保护图像边缘细节,既初步削弱高斯噪声又保护了边缘,最后采用改进的小波阈值滤波算法,对不同的小波系数采用不同的阈值函数,通过线性回归得到各最优阈值关系式。实验结果表明,该混合噪声自适应滤除算法能有效滤除椒盐噪声和高斯噪声,在图像主观质量和客观质量上均取得了较好的效果,能提高去噪图像峰值信噪比0.5~2.0 dB。