Papers by Keyword: Wavelet Transform (WT)

Abstract: In speech signal processing, the techniques of speech segmentation as front end of preprocessing have great importance in speech enhancing, coding and recognition. This paper analyzes the performances of several typical algorithms of speech segmentation, which are compared with each other. It put emphasis on the study of the algorithm based on the wavelet transformation. The smooth and gradual changing low frequency component can not segment the speech efficiently. In order to solve the problem, this paper put forward to an algorithm based on the cumulate energy of the wavelet transformation which promotes the precision of the segmentation on the phoneme level. But as a result of the wavelet sensitivity, it will present certain number of false spots. Therefore this paper proposes tow methods removing false spots. Finally it makes certain summary to these technologies.

Abstract: Tool condition monitoring is an important issue in the advanced machining process. Existing methods of tool wear monitoring is hardly suitable for mass production of cutting parameters fluctuation. In this paper, a new method for milling tool wear condition monitoring base on tunable Q-factor wavelet transform and Shannon entropy is presented. Spindle motor current signals were recorded during the face milling process. The wavelet energy entropy of the current signals carries information about the change of energy distribution associated with different tool wear conditions. Experiment results showed that the new method could successfully extract significant signature from the spindle-motor current signals to effectively estimate tool wear condition during face milling.

Abstract: The traditional identification method of surface crack has some shortcomings such as detection equipment complex, high requirements for operator and not suitable for large-scale structure inspection, vibration method is applied to identify surface crack in steel rods in this paper, using the method of calculating modal analysis combined with time-frequency wavelet analysis. The research on damage identification is performed on steel rods with different deep crack. The research results show that the accurate damage location can be judged by wavelet modulus maximum and the quantitative analysis of single crack identification can be achieved by constructing mathematical expressions of a single crack damage degree and damage index Lipschitz index. The research would province some guidance for engineering applications.

Abstract: Based on wavelet threshold de-noising method which put forward by Donoho, we analyze and compare the advantages and disadvantages of hard threshold, soft threshold and some improved threshold methods. Based on polynomial interpolation method, a new threshold function is proposed, which is continuous in the whole threshold area, with little constant deviation between the estimated wavelet coefficients and original signal wavelet coefficient, and high-order derivative and easy processing. The simulation results show that this method can give better optical effect, SNR gains and MSE performance.

Abstract: The measurement noise variance in the process of EKF is prone to bring error accumulation and can lead to filter divergence. Aiming at this kind of shortcoming, in this paper we build model of target motion observed on a single measurement point in a two-dimensional plane firstly. Secondly, we compare two methods, the variance estimation based on the signal-to-noise separation of wavelet transform and EKF algorithm based on noise variance estimation, applying in target tracking. Then, we adopt the wavelet transform to distinguish noise from the measurement signal real-timely. And the median variance estimator is used to estimate the measurement noise, which can improve the precision in EKF of target tracking by combining with EKF. Finally, the method of Monte Carlo simulation is used to prove its effectiveness and practicality.

Abstract: This paper introduces the related theory of sound waves , we collect the data through the acoustic wave sensors of the pipeline fault diagnosis system platform, decompose the signals to five layers by Mallat algorithm and wavelet function db4, compare the normal waves and leakage acoustic signal spectrum, and then get the power spectrum estimation for the decomposed signal at each level, we can see the signals energy feature in different frequency band. Feature extraction method based on wavelet transform can make the category of signal characteristics fully displayed in the different resolution band, it has a good application prospect in the field of acoustic signal processing.

Abstract: Magnetic anomaly detection is a passive method for detection of a ferromagnetic target, and its performance is often limited by external noise with a power spectral density of 1/f ^a, (0<a<2). In consideration of this kind of noise is non-stationary, self-similarity and long-range correlation, an effective noise reduction method based on the wavelet transform is proposed in this paper. The proposed method is only take one parameter into account, while the hard thresholding and soft thresholding methods utilize the relationship of the variance of the noisy signal. The simulation results show that the performance of our proposed method is superior to that of other methods.

Abstract: In this paper the application of wavelet in data detection of dynamic testing is chiefly researched , i.e. , by applying the method of wavelet denoising to eliminate the non-stationary random noise which produced in data detection of dynamic testing , by analyzing dynamic testing data to define the optimal wavelet as well as the optimal decomposition scale. Based on actual requirements of dynamic testing system, to reconstruct the data accurately by using FIR filter with biorthogonal wavelet, the method has a favorable effect.

Abstract: The image definition identification method based on the composite model of wavelet transform and neural networks is stronger in image edge character extraction, nonlinear process, self-adapted study and pattern recognition. The paper puts forward an evaluation method of image definition based on the focusing mechanism of simulating persons eyes by neural networks and on the composite model of wavelet transformation and neural networks. The wavelet component statistics obtained by the wavelet transform are taken as the inputs of the 5 layer RBF neural network model. The model identifies the image definition applying the steepest descent method of the additional momentum in a variable step size to adjust the network weights. The compound model is first trained by 75 images from the training set, and then is tested by 102 images from the testing set. The results show that this is a very effective identification method which can obtain a higher recognition rate.

Abstract: In this paper, a multi-resolution wavelet threshold denoising method which can achieve radar weak single detection is proposed. The method and the way of the radar weak signal detection which based on the Wavelets transform are described. The simulation to real radar signal is verified by MATLAB. The good results are obtained that the proposed method effectively improves the signal noise rate of the weak signal detection.