小波包自适应阈值语音降噪新算法

为了克服低信噪比输入下,语音增强造成语音清音中的弱分量损失,造成重构信号包络失真的问题。论文提出了一种新的语音增强方法。该方法根据语音感知模型,采用不完全小波包分解拟合语音临界频带,并对语音按子带能量进行清浊音区分处理,在阈值计算上,提出了一种清浊音分离,基于子带信号能量的小波包自适应阈值算法。通过仿真实验,客观评测和听音测试表明,该算法在低信噪比输入时较传统算法,能够更加有效地减少重构信号包络失真,在不损伤语音清晰度和自然度的前提下,使输出信噪比明显提高。将该算法与能量谱减法结合,进行二次增强能进一步提高降噪输出的语音质量。     

Algorithm for Detecting a Sequence of Pulsed Signals with Random Appearance Times

Using the methods of optimal nonlinear Markov filtering, we obtain an algorithm for optimal detection of a sequence of pulsed signals with random times of appearance against the background of white Gaussian noise in discrete time. Statistical characteristics of the synthesized algorithm are studied by computer simulation. The dependences of correct-detection probability on the signal-to-noise ratio and the intervals between pulses are obtained. It is shown that using the optimal nonlinear filtering methods we can improve the quality of detection of a sequence of pulsed signals compared with the methods based on the matched filtering of individual pulses in a packet.     

Noise reduction in ultrasonic NDT using undecimated wavelet transforms

Translation-invariant wavelet processing is applied to grain noise reduction in ultrasonic non-destructive testing of materials. In particular, the undecimated wavelet transform (UWT), which is essentially a discrete wavelet transform (DWT) that avoids decimation, is used. Two different UWT processors have been specifically developed for that purpose, based on two UWT implementation schemes: the "à trous" algorithm and the cycle-spinning scheme. The performance of these two UWT processors is compared with that of a classical DWT processor, by using synthetic grain noise registers and experimental pulse-echo NDT traces. The synthetic ultrasonic traces have been generated by an own-developed frequency-domain model that includes frequency dependence in both material attenuation and scattering. The experimental ultrasonic traces have been obtained by inspecting a piece of carbon-fiber reinforced plastic composite in which we have mechanized artificial flaws. Decomposition level-dependent thresholds, which are suitable for correlated noise, are specifically determined in all cases. Soft thresholding, Daubechies db6 mother wavelet and the three well-known threshold selection rules, Universal, Minimax and SURE, are applied to the different decomposition levels. The performance of the different de-noising procedures for single echo detection has been comparatively evaluated in terms of signal-to-noise ratio enhancement.     

Perceptually motivated wavelet packet transform for bioacoustic signal enhancement

A significant and often unavoidable problem in bioacoustic signal processing is the presence of background noise due to an adverse recording environment. This paper proposes a new bioacoustic signal enhancement technique which can be used on a wide range of species. The technique is based on a perceptually scaled wavelet packet decomposition using a species-specific Greenwood scale function. Spectral estimation techniques, similar to those used for human speech enhancement, are used for estimation of clean signal wavelet coefficients under an additive noise model. The new approach is compared to several other techniques, including basic bandpass filtering as well as classical speech enhancement methods such as spectral subtraction, Wiener filtering, and Ephraim-Malah filtering. Vocalizations recorded from several species are used for evaluation, including the ortolan bunting (Emberiza hortulana), rhesus monkey (Macaca mulatta), and humpback whale (Megaptera novaeanglia), with both additive white Gaussian noise and environment recording noise added across a range of signal-to-noise ratios (SNRs). Results, measured by both SNR and segmental SNR of the enhanced wave forms, indicate that the proposed method outperforms other approaches for a wide range of noise conditions.     

Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics

De-noising and extraction of the weak signature are crucial to fault prognostics in which case features are often very weak and masked by noise. The wavelet transform has been widely used in signal de-noising due to its extraordinary time-frequency representation capability. In this paper, the performance of wavelet decomposition-based de-noising and wavelet filter-based de-noising methods are compared based on signals from mechanical defects. The comparison result reveals that wavelet filter is more suitable and reliable to detect a weak signature of mechanical impulse-like defect signals, whereas the wavelet decomposition de-noising method can achieve satisfactory results on smooth signal detection. In order to select optimal parameters for the wavelet filter, a two-step optimization process is proposed. Minimal Shannon entropy is used to optimize the Morlet wavelet shape factor. A periodicity detection method based on singular value decomposition (SVD) is used to choose the appropriate scale for the wavelet transform. The signal de-noising results from both simulated signals and experimental data are presented and both support the proposed method.     

Wavelet based noise suppression technique and its application to ultrasonic flaw detection

The noise suppression techniques with wavelet transform (WT) are widely used in non-destructive testing and evaluation (NDT&E), especially in ultrasonics. Complete reconstruction theory with hard or soft thresholds, reconstruction technique based on the singularities of noise and signal, matched filter with an impulse response, and optimal frequency-to-bandwidth ratio of wavelet technique have all been used to analyze ultrasonic signals for noise suppression. But a more simple and effective technique has been pursued for decades. This paper develops a new technique using WT for the right purpose. In this work, WT is treated as a band-pass filter whose central frequency and frequency bandwidth (CF&FB) are determined by the spectra distribution of an ultrasonic signal captured from real testing situation. For the purpose of matching their CF&FB well, a technique for evaluating the optimal scale of a daughter wavelet is carried out too. By acting this daughter wavelet as a band-pass filter, we can obtain excellent de-noising results, even when the signal to noise ratio (SNR) is below -18 dB. The performance of the technique has been done by ultrasonic signals with computer generated white noises. Finally, the experimental verification is performed on a pipeline specimen with man-made small flaws with good results obtained. The results show that the technique is more suitable for processing heavy noised ultrasonic signals, and it can also be used in automatic flaw detection.     

基于小波变换的紫外光信号降噪处理

介绍了小波变换的基本理论以及利用小波变换进行信号处理的方法和步骤.选择db8小波对紫外光信号进行小波阈值降噪处理.实验结果表明采用小波阈值降噪法能够有效地抑制信号中的噪声,信噪比提高了8.519 2 dB.     

基于小波变换的信号去噪在瑞利-泰勒不稳定性定标实验中的应用

 采用Daubechies小波分析,对神光Ⅱ装置上流体力学不稳定性实验中样品扰动幅度和X光强度之间定标关系的数据进行了处理,获得了比较理想的去噪效果,并由此得到了平面调制靶波长及材料线性吸收系数等细节信息。该定标关系对准确测量瑞利-泰勒不稳定性的增长至关重要。作为对比,利用Wiener滤波方法对数据进行了处理,结果显示在处理这类信号时,Daubechies小波滤波在去除噪声和保留信号细节特征方面明显优于Wiener滤波。     

基于小波变换的低场核磁共振信号去噪方法的研究进展

低场核磁共振（low-field Nuclear Magnetic Resonance,low-field NMR）技术因其自身具有的独特优越性常被应用于极端条件下的测量,而且由于其采用的是永磁体,因而采集到的信号信噪比常常较低,在很大程度上影响了测量值的准确性.因此,如何去除混杂在信号中的加性高斯白噪声增加测量值的可靠性显得尤为重要.针对这一问题,国内外学者相继提出了众多优秀的去噪方法,其核心都是在不损失含噪信号中有效信息的基础上滤除掉夹杂在其中的噪声信号.本文在基于对小波变换理论分析的基础上,介绍了3种目前较为流行的用于低场核磁共振信号去噪的方法,分别是小波阈值去噪、模极大值去噪和小波系数相关性去噪,并给出了用于评价去噪效果的四个参数及其计算方法.     

Speech enhancement based on the discrete Gabor transform and multi-notch adaptive digital filters

This paper presents a new method to speech enhancement based on time-frequency analysis and adaptive digital filtering. The proposed method for dual-channel speech enhancement was developed by tracking frequencies of corrupting signal by the discrete Gabor transform (DGT) and implementing multi-notch adaptive digital filter (MNADF) at those frequencies. Since no a priori knowledge of the noise source statistics is required this method differs from traditional speech enhancement methods. Specifically, the proposed method was applied to the case where speech quality and intelligibility deteriorate in the presence of background noise. Speech coders and automatic speech recognition (ASR) systems are designed to act on clean speech signals. Therefore, corrupted speech signals by the noise must be enhanced before their processing. The method uses a primary input containing the corrupted speech signal while a reference input containing the noise only. In this paper, we designed MNADF instead of single-notch adaptive digital filter and used DGT to track frequencies of corrupting signal because fast filtering process and fast measure of the time-dependent noise frequency are of great importance in speech enhancement process. Therefore, MNADF was implemented to take advantage of fast filtering process. Different types of noises from Noisex-92 database were used to degrade real speech signals. Objective measures, the study of the speech spectrograms and global signal-to-noise ratio (SNR), segmental SNR (segSNR), Itakura-Saito distance measure as well as subjective listing test demonstrated consistently superior enhancement performance of the proposed method over traditional speech enhancement method such as spectral subtraction. Combining MNADF and DGT, excellent speech enhancement was obtained.     

Influence of thresholding procedures in ultrasonic grain noise reduction using wavelets

Wavelet transform based techniques are used for signal-to-noise ratio (SNR) enhancement in ultrasonic non-destructive testing and evaluation of strong sound scattering materials. The overall denoising performance of a wavelet signal processor is conditioned by several processing parameters, including the type of wavelet, thresholding method, and threshold selection rules. Different thresholding procedures and threshold selection rules are analysed in this paper using the discrete wavelet transform and decomposition level dependent thresholds. Global performance is evaluated by means of the SNR enhancement using synthetic grain noise registers with an incrusted flaw signal, with different values of the input SNR, and experimental ultrasonic traces acquired from a carbon fibre reinforced plastic composite block.     

基于双高斯衰减模型的超声回波处理方法

信号降噪与特征提取是超声检测数据处理的关键技术.基于超声信号有特定结构而噪声和超声信号的结构无关,本文提出一种旨在解决强噪声背景下超声回波的参数估计和降噪问题的方法.该方法将超声回波的参数估计和降噪问题转换为函数优化问题,首先根据工程经验建立超声信号的双高斯衰减数学模型,然后根据观测回波和建立的超声信号模型确定目标函数,接着选择人工蜂群算法对目标函数进行优化从而得到参数的最优估计值,最后由估计出的参数根据建立的超声信号数学模型重构出无噪的超声估计信号.通过仿真和实验表明本文方法可以准确估计出信噪比大于-10 dB的含噪超声回波中的无噪信号,且效果优于基于自适应阈值的小波降噪方法和经验模态分解方法;此外相比常用的指数模型和高斯模型,本文提出的双高斯衰减超声信号模型与实测超声信号更接近,其均方误差为9.4×10~(-5),波形相似系数为0.98.     

Some optimizations on detecting gravitational wave using convolutional neural network

This work investigates the problem of detecting gravitational wave (GW) events based on simulated damped sinusoid signals contaminated with white Gaussian noise. It is treated as a classification problem with one class for the interesting events. The proposed scheme consists of the following two successive steps: decomposing the data using a wavelet packet, representing the GW signal and noise using the derived decomposition coefficients; and determining the existence of any GW event using a convolutional neural network (CNN) with a logistic regression output layer. The characteristic of this work is its comprehensive investigations on CNN structure, detection window width, data resolution, wavelet packet decomposition and detection window overlap scheme. Extensive simulation experiments show excellent performances for reliable detection of signals with a range of GW model parameters and signal-to-noise ratios. While we use a simple waveform model in this study, we expect the method to be particularly valuable when the potential GW shapes are too complex to be characterized with a template bank.     

Ultrasonic flaw detection in NDE of highly scattering materials using wavelet and Wigner-Ville transform processing

In ultrasonic non-destructive evaluation of highly scattering materials the backscattering noise may attain peak values greater than the searched flaw pulse and the mean value of noise spectrum is very similar to the searched echo spectrum. Several specific methods have been proposed for the reduction of this type of noise, but the comparison of the performance of different methods is still an open problem. In this paper, we make a comparison among some methods based on simultaneous representations in time and frequency/scale domains of the ultrasonic traces. Synthetic and experimental traces are de-noised using a discrete wavelet processor with decomposition level-dependent threshold selection and a method that combines Wigner-Ville transform and filtering in the time-frequency domain. The results are comparatively evaluated in terms of signal to noise ratio and probability of detection.     

结点阈值小波包变换图像去噪新算法

小波包变换是小波变换的推广,可视为普通小波函数的线性组合,具有灵活的时频分析能力,随着分解层数的增加,小波包分解能够在所有的频率范围聚焦。提出一种应用结点阈值小波包变换的新型图像去噪算法。利用小波包变换对含噪图像进行分解,在图像信号的子带层次上进行结点阈值操作,采用软阈值的方法进行阈值处理,结点噪声采用谱熵法估计,并使用峰值信噪比评估去噪后的图像质量。实验结果表明,相比于使用其它阈值方法的小波包图像去噪算法,该算法具有更好的图像去噪性能。     

Extraction of weak crack signals based on sparse code shrinkage combined with wavelet packet filtering

Early crack signals in critical infrastructure components of major equipment are hardly to be extracted due to its low signal noise ratio (SNR). A de-noising method combined wavelet packet (WP) technology with sparse code shrinkage (SCS) is proposed in this study. Firstly, WP reconstruction technology is used to reserve the crack signal with a specified frequency range. That is, the signal is decomposed by Meyer wavelet into five layers, and the signal with the frequency range from 187.5 kHz to 609.375 kHz is reserved. Then SCS method removes noise within the specified frequency range. Namely, the probability density function (PDF) of the signal independent coefficients is estimated via the generalized Gaussian model (GGM) in the independent component analysis (ICA) space. The nonlinear de-noising is finished by utilizing maximum a posteriori (MAP) estimate. The results obtained by the combined method are compared with those generated by the SCS method and the WP de-noising method. It demonstrates that the combined method is the best one among the three methods in extracting weak signals. Its output SNR is −2.38 dB and the correlation coefficient (CC) is 0.54 when the input SNR is −20 dB. They are higher than those obtained by the SCS method (SNR −4.46 dB and CC 0.51). The WP method is the worst (SNR −3.54 dB and CC −0.003). Therefore, the combined method is quite suitable for weak signal extraction.     

一种强噪声背景下微弱超声信号提取方法研究

为解决在强噪声背景下获取超声信号的难题,基于粒子群优化算法和稀疏分解理论提出一种强噪声背景下微弱超声信号提取方法.该方法将降噪问题转换为在无穷大参数集上对函数进行优化的问题,首先以稀疏分解理论和超声信号的结构特点为依据构建了粒子群优化算法运行所需要的目标函数及去噪后信号的重构函数,从而将粒子群优化算法和超声信号降噪联系在一起;然后根据粒子群优化算法可以在连续参数空间寻优的特点建立了用于匹配超声信号的连续超完备字典,并采用改进的自适应粒子群优化算法在该字典中对目标函数进行优化;最后根据对目标函数在字典上的优化结果确定最优原子,并利用最优原子按照重构函数重构出降噪后的超声信号.通过对仿真超声信号和实测超声信号的处理,结果表明本文提出的方法可以有效提取信噪比低至-4 dB的强噪声背景下的微弱超声信号,且和基于自适应阈值的小波方法相比本文方法表现出更好的降噪性能.     

Signal enhancement of a novel multi-address coding lidar backscatters based on a combined technique of demodulation and wavelet de-noising

Multi-address coding (MAC) lidar is a novel lidar system recently developed by our laboratory. By applying a new combined technique of multi-address encoding, multiplexing and decoding, range resolution is effectively improved. In data processing, a signal enhancement method involving laser signal demodulation and wavelet de-noising in the downlink is proposed to improve the signal to noise ratio (SNR) of raw signal and the capability of remote application. In this paper, the working mechanism of MAC lidar is introduced and the implementation of encoding and decoding is also illustrated. We focus on the signal enhancement method and provide the mathematical model and analysis of an algorithm on the basis of the combined method of demodulation and wavelet de-noising. The experimental results and analysis demonstrate that the signal enhancement approach improves the SNR of raw data. Overall, compared with conventional lidar system, MAC lidar achieves a higher resolution and better de-noising performance in long-range detection.     

小波包节点分段阈值降噪在水声监听中的应用*

水声目标信号在发送、传播过程中,易受到环境噪声、系统自噪声等影响,因此水声监听过程中目标信号会掺杂大量噪声信息。为提高获取目标信号的准确性和可靠性,降低噪声,在已有小波分析基础上,提出小波包节点相对能量判断最优分解层,最优分解层节点系数分段阈值处理重构方法,实现水声监听信号分频段去噪。将0.1 kHz~8.4 k Hz实验数据按节点频率排序划分为5个强弱不同的频段信号实现消噪提取,结果表明该方法可将噪声信号与目标信号有效分离,与全局单一阈值相比,具有较好降噪能力。该方法打破了小波阈值去噪高频处理的局限性,提高了识别精度,改善了全局单一阈值去噪存在的短板,在鱼类分析识别、舰船监听、深海探测等方面具有一定的推广和应用价值。     

基于双尺度相关运算的拉曼谱峰识别方法

拉曼谱峰识别是拉曼定性分析中的关键技术之一, 针对现有拉曼谱峰识别方法中存在的缺陷和不足提出了一种双尺度相关拉曼光谱谱峰识别方法,即采用两个尺度下的相关系数与局部信噪比相结合来实现拉曼谱峰识别。利用MATLAB对所提算法与传统的连续小波变换法进行了对比分析,并通过实测拉曼光谱进行验证。分析结果：双尺度相关法识别一幅拉曼谱的平均时间为0.51 s,连续小波变换法为0.71 s;当谱峰信噪比≥6时(现代拉曼光谱仪器均可达到较高的信噪比),双尺度相关法的谱峰识别准确率高于99%,连续小波变换法的谱峰识别准确率小于84%,且双尺度相关法谱峰位置识别误差的均值与标准差均要小于连续小波变换法。通过仿真对比分析和实验验证表明：双尺度相关法具有无需人工干预,无需做去噪及去背景等预处理操作,识别速度快,识别准确率高等特点,是一种切实可行的拉曼谱峰识别方法。