Papers by Keyword: Empirical Mode Decomposition (EMD)

Abstract: Electrocardiogram (ECG) is the most important signal in the biomedical field for the diagnosis of Cardiac Arrhythmia (CA). ECG signal often interrupted with various noises due to non-stationary nature which leads to poor diagnosis. Denoising process helps the physicians for accurate decision making in treatment. In many papers various noise elimination techniques are tried to enhance the signal quality. In this paper a novel hybrid denoising technique using EMD-DWT for the removal of various noises such as Additive White Gaussian Noise (AWGN), Baseline Wander (BW) noise, Power Line Interference (PLI) noise at various concentrations are compared to the conventional methods in terms of Root Mean Square Error (RSME), Signal to Noise Ratio (SNR), Peak Signal to Noise Ratio (PSNR), Cross-Correlation (CC) and Percent Root Square Difference (PRD). The average values of RMSE, SNR, PSNR, CC and PRD are 0.0890, 9.8821, 14.4464, 0.9872 and 10.9036 for the EMD approach, respectively, and 0.0707, 10.7181, 16.2824, 0.9874 and 10.7245 for the proposed EMD-DWT approach, respectively, by removing AWGN noise. Similarly BW noise and PLI are removed from the ECG signal by calculating the same quality metrics. The proposed methodology has lower RMSE and PRD values, higher SNR, PSNR and CC values than the conventional methods.

Abstract: Aiming at the mode mixing problem caused by interpolation point selection of conventional EMD (Empirical mode decomposition) method, a secondary iterative sifting EMD method that can avoid mode mixing and achieve high-precision decomposition of HHT (Hilbert–Huang transformation) is proposed based on the theory of EMD. The simulation results show that the proposed method is superior to conventional EMD on the ability to split mixed signal. Finally, the proposed algorithm is applied to the fault diagnosis of rolling bearing and the test results have proved its effectiveness and advantages.

Abstract: In this paper, considering the nonlinear and non-stationary properties of extreme high-temperature time series, we introduce Empirical Mode Decomposition to analyze the extreme high-temperature time series from 1959 to 2012 in Fengxian district of Shanghai. The scale characteristics and oscillating mode characteristics were mainly investigated. The trend of extreme high-temperature also shows periodic variation from decreasing to increasing for the recent fifty years. Analyze the reconstructed modes with the wave pattern: It shows that variability are quite large from 1997 to 1999 and from 1977 to 1982, which shows extreme high-temperature rose and fell dramatically in these periods. The volatility from 2006 to 2008 is far more dramatic than the other times. And it is the most remarkable in the recent fifty years.

Abstract: In order to improve the accuracy of the phase and amplitude of acceleration integration results, a new method to transform acceleration signal into displacement was presented which combines Empirical Modal Decomposition (EMD) adaptive filtering with FFT based frequency domain integration. The acceleration signal is decomposed by EMD into n IMF, and by certain rules, the number of IMF pertaining high frequency (h) is determined, and the h IMF are adaptively filtered to cancel noises. After that, the FFT transform is applied to the n processed IMF, frequency domain integration is done, and finally the displacements time series is obtained by IFFT. Simulation shows that this acceleration signal processing method is better than the pure frequency domain integration transformation.

Abstract: According to the non-stationary and non-linear characteristics of poultry voice and the situation that it`s hard to obtain enough sound samples, a poultry voice classification method based on Empirical Mode Decomposition (EMD), Teager energy transformation, and Support Vector Machine (SVM) is proposed. Firstly, the poultry voice signals are decomposed into a finite number of intrinsic mode function (IMF).Then, the Teager energy of five IMFs filtered are used to form characteristic vectors. Finally, the eigenvectors are put into a support vector machine classifier . The results of animal voice signals experimental recognition showed that this method had high accuracy and good generalization abilities even in the case of small number of samples. The approach proposed could identify the poultry voice effectively.

Abstract: This paper puts forward a classification method of imitating the human eyes to recognize image as a whole which combined chaotic neural network and the Empirical Mode Decomposition (EMD). The method takes the individual of weeds plant as the research object and utilizes the chaotic neural network, the EMD, Teager energy operator and cluster analysis technology comprehensively. Firstly take the two-dimensional gray image matrix as the chaotic neural network weight matrix directly. Use the chaotic neural network with n neurons expression to iterate and get the one-dimensional output curve. Secondly, make the EMD decomposition for the curve and get the corresponding intrinsic mode function (IMF) curves. Then make the Teager energy transformation for each IMF component and get the average Teager energy. Finally use the fuzzy clustering algorithm to cluster analyze and get the clustering results. It can realize the classification of different categories of weeds through analysis and contrast the results with the original images. The experiment proves the effectiveness of the proposed algorithm for classifying weeds, and it is a universal new method of weeds classification.

Abstract: The Hilbert-Huang Transform (HHT) is a new time-frequency analysis with adaptability and orthogonality, but it is rather weak in terms of noise resistance, even low noise can disturb the HHT result greatly. The paper launches an investigation on how noises affect the HHT result and proposes the method to solve the problem. The analytic framework for HHT is first introduced, the feature of the test signal is extracted by HHT. Median filter is adopted to reduce the frequency leakage of certain signal component caused by white noise. The method proposed is experimentally simulated and the results demonstrate its effectiveness.

Abstract: Aiming at traditional methods cannot get good performance in noisy environments, an improved method for speech enhancement based on Empirical Mode Decomposition (EMD) and Morphology Filtering (MF) was proposed. The method firstly uses EMD to obtain Intrinsic Mode Function (IMF) and for hard threshold processing, then selects appropriate structuring element to construct MF for filtering processing in remaining IMFs. Finally, speech enhancement signal is reconstructed for each IMFs. Experimental results show that the proposed method for speech enhancement has better de-noising effect by comparing time-domain waveform and spectrogram. Moreover, the quality of reconstructed speech enhancement signal has been significantly improved.

Abstract: In non-parametric signal denoising area, empirical mode decomposition is potentially useful. In this paper, the wavelet thresholding principle is directly used in EMD-based denoising. The basic principle of the method is to reconstruct the signal with IMFs previously thresholded. A novel threshold function is proposed to improve denoising effect by exploiting the special characteristics of the hard and soft thresholding method. The denoising method is validated through experiments on the “Doppler” signal and a real ECG signal from MIT-BIH databases corrupted by additive white Gaussian random noise. The simulations show that the proposed EMD-based method provides very good results for denoising.

Abstract: Based on the principal of forecast of Artificial Neural Network, Radial Basis Function neural network and Radial Basis Function neural network based on EMD were introduced into the field of precipitation forecasting in this article. With the precipitation data of 27 sites from1950-2010, EMD-RBF network was set up, and the difference between the predictive value and the actual precipitation data was discussed. The results showed that the correlation Of EMD-RBF forecast precipitation and actual precipitation is more than 0.9. Of all sites, the maximum relative prediction error of 17 sites is less than 10%, the maximum relative error does not exceed 15%.The EMD-RBF model had good quality on forecasting precision, which provided a new method for precipitation forecasting.