基于三次样条重建的超声多普勒血流信号提取

根据超声多普勒血流信号和血管壁搏动信号的不同统计特性,对采集的多普勒信号进行非均匀抽样和三次样条函数重建,从中得到血管壁搏动信号,从而提取出超声多普勒系统中的血流信号.对计算机仿真的超声多普勒信号和采集的人体颈总动脉多普勒信号分别应用该方法,并和传统的高通滤波器方法进行比较.实验结果表明:本方法能在滤除管壁搏动信号的同时,保留低频部分的超声多普勒血流信号,可用于超声多普勒系统中两种信号成份的分离.     

总体经验模态细分法提取血流超声多普勒信号的研究

提出基于总体经验模态分解（EEMD）血流细分法提高血流超声多普勒信号提取精度.首先估计辅助分析所需的白噪声幅度,进而用EEMD得到无模态混叠的本征模态函数（IMF）组,最后分离出血流信号的IMF.将本方法应用于计算机仿真和人体实测超声多普勒信号,并与高通滤波器法、原EMD法和EMD细分法比较.结果表明本文方法,提取的血流信号精度最高,特别对WBSR=70dB的混合信号,其精度比上述方法分别提高35%、38%及17%.     

A new clutter rejection algorithm for Doppler ultrasound

Several strategies, known as clutter or wall Doppler filtering, were proposed to remove the strong echoes produced by stationary or slow moving tissue structures from the Doppler blood flow signal. In this study, the matching pursuit (MP) method is proposed to remove clutter components. The MP method decomposes the Doppler signal into wavelet atoms that are selected in a decreasing energy order. Thus, the high-energy clutter components are extracted first. In the present study, the pulsatile Doppler signal s(n) was simulated by a sum of random-phase sinusoids. Two types of high-amplitude clutter signals were then superimposed on s(n): time-varying low-frequency components, covering systole and early diastole, and short transient clutter signals, distributed within the whole cardiac cycle. The Doppler signals were modeled with the MP method and the most dominant atoms were subtracted from the time-domain signal s(n) until the signal-to-clutter (S/C) ratio reached a maximum. For the low-frequency clutter signal, the improvement in S/C ratio was 19.0 +/- 0.6 dB, and 72.0 +/- 4.5 atoms were required to reach this performance. For the transient clutter signal, ten atoms were required and the maximum improvement in S/C ratio was 5.5 +/- 0.5 dB. The performance of the MP method was also tested on real data recorded over the common carotid artery of a normal subject. Removing 15 atoms significantly improved the appearance of the Doppler sonogram contaminated with low-frequency clutter. Many more atoms (over 200) were required to remove transient clutter components. These results suggest the possibility of using this signal processing approach to implement clutter rejection filters on ultrasound commercial instruments.     

一种改进的经验模型分解方法

在对复杂信号进行分析中,常把它展开成一系列基本信号,然后,通过研究每个基本成分或者相应系数的特点来分析复杂信号。Huang等人提出经验模型分解方法(Empirical Mode Decomposition,EMD),通过筛选,将复杂信号中分解成一系列内在模型函数(Intrinsic Mode Function,IMF)。在本论文中,作者对经验模型分解中的一个重要的筛选过程作了部分改进,提出了一种改进检验模型分解法(Modified EMD,MEMD)。利用改进检验模型分解法,能够既快又准确地获得内在模型函数,而且,得到的内在模型函数能保留原信号中各成分的瞬时频率的规律。     

The effect of noise and high-pass filtering on the estimation ofmean blood velocity using wide and narrow ultrasound beams

Recently, a new method of analysis has been proposed for the calculation of a Doppler frequency proportional to mean blood velocity for the case where the Doppler beam is assumed to be of negligible thickness compared to the vessel diameter, and the velocity profile is axisymmetric and monotonic increasing from the vessel wall to the vessel center. Such analysis of the Doppler signal is an alternative to that commonly performed under the assumption that the beam insonates the vessel uniformly. Errors in each method are found and compared for the case where the Doppler signal is contaminated by noise, and for the case where the signal is subjected to an ideal high-pass filter. The frequency resulting from the new method of analysis is affected by low-frequency perturbations approximately twice as much as that resulting from the standard method. However, the new method is much more immune to high frequency perturbations. If each method is used with the beam shape for which it is appropriate then, for a given velocity profile, each method is equally affected by the use of the same ideal high-pass filter     

An Analytical Expression for Empirical Mode Decomposition Based on B-Spline Interpolation

Although empirical mode decomposition (EMD) lacks a rigorous theoretical basis, it has attracted much attention for analyzing nonstationary signals adaptively. In this paper, the EMD method is investigated from a digital signal processing perspective. Based on an analysis of extrema sampling and B-spline interpolation, we show that the upper and lower envelopes of signals are formed by a succession of three basic operations: decimation of local extrema, interpolation, and filtering by a B-spline filter. We then show that some aliasing noise can be suppressed by the mean of the envelopes, though the extrema sampling is a sub-Nyquist sampling. For uniformly spaced extrema of signals, we derive a general analytical expression of intrinsic mode functions (IMFs) extracted by the EMD method from signals.     

高分七号激光测高中全波形回波数据的EMD降噪

针对具有多个高度层的复杂场景,全波形激光测高系统记录的回波信号中往往带有较高的噪声,采用合适的降噪方法将有助于提高计算激光测距的精确性、反演地物垂直结构和构建目标特征参数的准确性。根据高分七号激光测高在轨探测的低信噪比全波形数据的特性,采用经验模态分解（Empirical mode decomposition,EMD）方法来构建典型的本征模函数（Intrinsic mode function, IMF）,对于分解出多个不同尺度IMF的筛选,比较了使用去除高频分量,阈值选取、Wavelet选取和去趋势波动分析（Detrended fluctuation analysis, DFA）等方法与策略,通过降噪效果及定量评价,测试结果表明EMD-DFA1与EMD-1IMF对高分七号激光测高的全波形数据具有较好的降噪效果,其次为EMD-Wavelet和EMD-Threshold。另外通过EMD-DFA1对单个波峰、混叠波峰、多个波峰等不同情况的全波形数据测试,结果表明该方法具有较好的自适应性。     

Method for eliminating mode mixing of empirical mode decomposition based on the revised blind source separation

Since mode mixing of empirical mode decomposition (EMD) is mainly caused by the intermittence and noise, we propose a novel method to eliminate mode mixing of EMD based on the revised blind source separation. To this aim, an optimal morphological filter is employed to eliminate the noise. As a result, the component of mode mixing caused by noise is suppressed. Furthermore, the de-noised signal is decomposed into different intrinsic mode function (IMF) components through the EMD algorithm. Since it is impossible to apply blind source separation to a single channel signal directly, the IMF component, which has mode mixing is chosen and reconstructed in the phase space. Following that, the equivalent hypothetical signals are obtained. Finally, an improved fixed-point algorithm based on independent component analysis (ICA) is introduced to separate the overlapping components. The analysis of simulation and practical application demonstrates that our proposed method can effectively tackle the mode mixing problem of EMD.     

基于CEEMD和排列熵的SAR窄带干扰抑制方法

为了有效抑制合成孔径雷达（SAR）系统中常见的窄带干扰（NBI）,本文提出一种基于互补集合经验模态分解（CEEMD）和排列熵（PE）的NBI抑制方法。矩峰度系数法用于检测原始回波中是否存在NBI,对包含NBI的回波使用CEEMD将其分解为一系列本征模态函数（IMF）。计算所有IMF排列熵得到全局阈值以区分NBI和有用信号,并使用去除NBI后的IMF分量重建信号以获得良好聚焦的SAR图像。结果表明：所提方法能有效克服经验模态分解（EMD）带来的模态混叠问题,且干扰抑制性能优于传统频域陷波法及基于EMD的NBI抑制方法。     

一种改进的基于EMD分解的超宽带信号消噪算法

针对低信噪比超宽带信号的消噪问题,提出一种改进的基于经验模式分解(EMD)的消噪算法.该算法首先对含噪信号进行EMD分解,得到多个固有模态函数(IMF)分量,然后选取高阶IMF重构原信号,达到消噪的目的.针对对UWB信号的IMF重构过程中阶数阈值难以确定的问题,通过数值仿真的方法,得到信号分量和噪声分量在不同阶IMF上的能量分布特性;在对所得特性进行分析的基础上,设计了一种数据自适应的阶数阈值选取算法,解决了EMD消噪中的阶数阈值选取问题.仿真结果表明,EMD消噪算法能够在较低信噪比下提供平均10 dB的信噪比增益,可以有效地对超宽带信号进行消噪.     

基于Hankel-SVD的非平稳超声血流成像杂波抑制技术研究

有效抑制由血管或血管周围组织时变运动引起的非平稳杂波对于提高诊断超声彩色血流成像中血流动力学参数描述的准确性有着极其重要的意义。该文基于奇异值滤波技术提出一种改进的非平稳杂波自适应抑制方法。该方法逐次利用单个慢时多普勒回波采样矢量构建Hankel矩阵,然后根据奇异值分解后得到的正交Hankel主成份所代表的频域内容,动态选取高阶Hankel主成份重构多普勒血流信号,实现非平稳杂波的有效抑制。为验证算法的有效性,分别对多普勒回波仿真模型合成数据与利用彩色超声设备(Sonix RP)采集的颈动脉血流基带回波信号进行滤波处理,然后采用滞一自相关估计法计算血流平均速度与功率并进行成像。处理结果表明,相对于传统IIR滤波方法与多项式回归滤波技术,利用该文所提算法可对高强度、非平稳杂波进行充分抑制,提高血流估计精度,此外,该算法具有空间自适应性,无需人为设定阈值参数以估计杂波空间维数,与现有基于特征分解的自适应滤波方法相比,可以有效提高组织空间高强度时变运动时血流与组织的区分能力。     

Speech Endpoint Detection in Noisy Environments Using EMD and Teager Energy Operator

Accurate endpoint detection is a necessary capability for speech recognition.A new energy measure method based on the empirical mode decomposition(EMD)algorithm and Tcager energy operator(TEO)is proposed to locate endpoint intervals of a speech signal embedded in noise.With the EMD,the noise signals can be decomposed into different numbers of sub-signals called intrinsic mode functions(IMFs),which is a zero-mean AM-FM component.Then TEO can be used to extract the desired feature of the modulation energy for IMF components.In order to show the effectiveness of the proposed method,examples are presented to show that the new measure is more effective than traditional measures.The present experimental results show that the measure can be used to improve the performance of endpoint detection algorithms and the accuracy of this algorithm is quite satisfactory and acceptable.     

基于主成分分析的经验模态分解消噪方法

 针对非线性非平稳信号的去噪问题,提出一种基于主成分分析(PCA)的经验模态分解(EMD)消噪方法.该方法根据EMD的分解特性,利用PCA对噪声信号经EMD分解后的内蕴模态函数(IMF)进行去噪处理:首先利用"3σ法则"对第一层IMF进行细节信息提取,并估计每层IMF中所含噪声的能量;然后对IMF进行PCA变换,根据IMF中所含噪声的能量选择合适数目的主成分分量进行重构,以去除IMF中的噪声.为验证本文方法的有效性,进行了数字仿真与实例应用实验.实验结果均表明,所提方法的消噪效果整体上优于Bayesian小波阈值消噪方法和基于模态单元的EMD阈值消噪方法,是一种有效的信号消噪新方法.     

A robust method for parameter estimation of AR systems using empirical mode decomposition

This paper presents a robust algorithm for parameter estimation of autoregressive (AR) systems in noise using empirical mode decomposition (EMD) method. The basic idea is to represent the autocorrelation function of the noise-free AR signal as the summation of damped sinusoidal functions and use EMD for extracting these component functions as intrinsic mode functions (IMFs). Unlike conventional correlation-based techniques, the proposed scheme first estimates the damped sinusoidal model parameters from the IMFs of autocorrelation function using a least-squares based method. The AR parameters are then directly obtained from the extracted sinusoidal model parameters. Simulation results show that EMD is a very promising tool for AR system identification at a very low signal-to-noise ratio (SNR).     

极点均值型经验模式分解及其去噪应用

使用经验模式分解（EMD）对信号进行去噪时,由于EMD 本身会产生模态混叠,往往很难将噪声完全分离。针对这一问题,提出了一种新型的极点均值型EMD 方法,并且给予固有模态函数（IMF）一个新的定义。首先,将相邻极点平均以求得均值包络,然后迭代相减进而获得IMF。最后用原始信号减去分离出的高频IMF 实现去噪。随机信号仿真以及激光雷达回波信号去噪实验表明,该方法与EMD 分解相比,可以更好地将噪声分离,有效地抑制模态混叠,更可以极大地减小均方误差。因此,极点均值型EMD 拥有很好前景。     

On the use of EMD based adaptive filtering for OFDM channel estimation

Hilbert Huang transform (HHT) based data driven empirical mode decomposition (EMD) in conjunction with adaptive filter (AF) is proposed for estimation of communication channel in OFDM system. EMD can be viewed as alike of wavelet decomposition which decomposes the signal of interest to intrinsic mode functions (IMF), whose basis function is derived from signal itself. In this method, the length of channel impulse response (CIR), is approximated using Akaike information criterion (AIC). Then the estimation of CIR is performed using adaptive filter with EMD decomposed IMF of the received OFDM symbol. Conventional AF uses random initial weight vector. The novelty of the proposed method lies in the fact that it uses decimated version of one of the decomposed IMFs of received OFDM symbol as initial weight vector. The selection of useful IMF component is done based on correlation and kurtosis measures. This makes the proposed EMD based AF method converge to minimum mean square error (MMSE) in less number of iterations resulting in almost 50% saving of computations. Bit error rate (BER), mean square error (MSE) and normalized root mean square error (NRMSE) are computed. The simulation studies established the efficacy of proposed method; and comparative studies under different modulation schemes and fading conditions revealed improved performance. Simulations have shown an average improvement of 3 dB in BER performance for proposed EMD based AF as compared to conventional AF.     

基于经验模态分解的模态域MVDR方法研究

 矢量阵MVDR波束形成可有效地实现信号的空间谱估计,但它仅适用于窄带信号,当各目标强度相差较大时,难以实现对弱目标的有效检测.经验模态分解具有突出信号局瞬特征的特点,可将多分量信号分解成多阶固有模态函数.结合固有模态函数特性和MVDR窄带信号要求,提出了矢量阵模态域MVDR波束形成算法,并将中心频率的概念应用于固有模态函数,以此作为模态域MVDR波束形成算法的中心频率.海试结果表明:本方法可增强弱目标所在方位空间谱的能量,有效地实现强干扰下弱目标的检测.     

基于CM-OMEMD和小波散射网络的语音情感识别

语音情感识别（Speech Emotion Recognition,SER）是人机交互的重要组成部分,具有广泛的研究和应用价值。针对当前SER中仍然存在着缺乏大规模语音情感数据集和语音情感特征的低鲁棒性而导致的语音情感识别准确率低等问题,提出了一种基于改进的经验模态分解方法（Empirical Mode Decomposition,EMD）和小波散射网络（Wavelet Scattering Network,WSN）的语音情感识别方法。首先,针对用于语音信号时频分析的EMD及其改进算法中存在的模态混叠问题（Mode Mixing）和噪声残余问题,提出了基于常数Q变换（Constant-Q Transform,CQT）和海洋捕食者算法（Marine Predator Algorithm,MPA）的优化掩模经验模态分解方法（Optimized Masking EMD based on CQT and MPA,CM-OMEMD）。采用CM-OMEMD算法对情感语音信号进行分解,得到固有模态函数（Intrinsic Mode Functions,IMFs）,并从IMFs中提取了可以表征情感的时频特征作为第一个特征集。然后采用WSN提取了具有平移不变性和形变稳定性的散射系数特征作为第二个特征集。最后将两个特征集进行融合,采用支持向量机（Support Vector Machine,SVM）分类器进行分类。通过在含有七种情感状态的TESS数据集中的对比实验,证明了本文提出的系统的有效性。其中CM-OMEMD减小了模态混叠,提升了对情感语音信号时频分析的准确性,同时提出的SER系统显著提高了情绪识别的性能。      

基于EMD时频重构特征的航管应答器个体识别

通过分析航管应答信号样式及特点,提出了基于经验模式分解（EMD）时频重构特征的航管应答器个体识别算法。首先通过EMD将多分量信号分解为有限个固有模式函数（IMF）分量,继而利用IMF来重构辐射源信号的时频分布,最终获得稳定的时频图分解特征。利用实测航管应答信号的验证实验表明,该方法可以有效提取航管应答信号的细微特征,最终的识别性能显著优于使用脉冲包络特征或者功率谱特征的识别算法。     

Gaussian Noise Filtering from ECG by Wiener Filter and Ensemble Empirical Mode Decomposition

Empirical mode decomposition (EMD) is a powerful algorithm that decomposes signals as a set of intrinsic mode function (IMF) based on the signal complexity. In this study, partial reconstruction of IMF acting as a filter was used for noise reduction in ECG. An improved algorithm, ensemble EMD (EEMD), was used for the first time to improve the noise-filtering performance, based on the mode-mixing reduction between near IMF scales. Both standard ECG templates derived from simulator and Arrhythmia ECG database were used as ECG signal, while Gaussian white noise was used as noise source. Mean square error (MSE) between the reconstructed ECG and original ECG was used as the filter performance indicator. FIR Wiener filter was also used to compare the filtering performance with EEMD. Experimental result showed that EEMD had better noise-filtering performance than EMD and FIR Wiener filter. The average MSE ratios of EEMD to EMD and FIR Wiener filter were 0.71 and 0.61, respectively. Thus, this study investigated an ECG noise-filtering procedure based on EEMD. Also, the optimal added noise power and trial number for EEMD was also examined.