Papers by Keyword: Modulus Maxima

Abstract: Wavelet transform denoising is an important application of wavelet analysis in signal and image processing. Several popular wavelet denoising methods are introduced including the Mallat forced denoising, the wavelet transform modulus maxima method and the nonlinear wavelet threshold denoising method. Their advantages and disadvantages are compared, which may be helpful in selecting the wavelet denoising methods. At the same time, several improvement methods are offered.

Abstract: In this paper, we introduce a texture image classification algorithm based on Gabor wavelet transform. Using Gabor wavelet transform, image is decomposed into sub-bands images in multiresolution and multi-direction, and we extract texture feature from all sub-bands images. Then the algorithm groups feature image into clusters by the k near neighbor algorithm. The experimental results on dataset Brodatz showed that the proposed algorithm can achieve an ideal accuracy rate and excellent classification effect.

Abstract: In ECG signals accurate detection to the position of QRS complex is a key to automatic analysis and diagnosis system. And its premise is that effectively remove all kinds of noise interference in ECG signal. Here, a method of detecting QRS based on EMD and wavelet transform was presented which is aim to improve the anti-noise performance of the detection algorithm. It is combined EMD with the theory of singularity detecting based on wavelet transform modulus maxima method. It has the high detection accuracy and good precision that can give an effective way to the automatic analysis for ECG signal.

Abstract: The basic principle of the wavelet analysis method to extract the mineral spectral absorption characteristics is summarized and different wavelet bases,wavelet decomposition scale extraction effect are analyzed in this paper. Then, the algorithm of mineral spectral absorption characteristics based on wavelet spectral analysis is provided and a large number of experiments are carried out using the spectrums of field rock mineral. Experimental result shows that such a method is less affected by original signal noise and depict the parameters of the mineral spectral absorption feature more precise, which is an ideal spectral feature extraction method compared with the traditional mineral spectral absorption feature extraction method.

Abstract: This paper presents a fusion algorithm for image edge detection based on the mathematical morphology and the NSCT. First the de-noised image is processed by the multi-structure elements of the mathematical morphology. And then the processed image is decomposed by the NSCT into multi-scale and multi-directional sub-bands. Edges in the high-frequency sub-bands are extracted with the dual-threshold modulus maxima method. Finally the edges of the de-noised image are refined into a single pixel edge image. The simulation results show that this method can effectively suppress noise, eliminate pseudo-edges, locate accurately and detect the complete outline.

Abstract: Dam safety monitoring data can be viewed as a digital signal sequence which consists of different frequency components. Identifying outliers to ensure the reliability of observational data becomes a foundation work of dam monitoring data analysis. Outliers of time signal series can be detected by wavelet transform. Lipschitz index can be used to measure the local singularity of a function, and the original abnormal signal can be found in the position of wavelet transform modulus maxima. Take horizontal displacement values observed by Lijiaxia concrete dam as example, an assumed error are added to the time series signal deliberately. A 4-level decomposition of the observation data was done by using wavelet db1, the results show that the modulus maxima occur at the given time. Therefore, outliers can be detected and located accurately by wavelet transform, which is important to analyze the safety monitoring data of dam.

Abstract: There are many outliers in air pollution time series data for various reasons. It has a serious impact on the data analysis and use. There are three main ways to identify anomalies but they each have definite limitations, especially when identifying and correcting the first category and the second category of outlier at the same time. In order to solve this problem, this paper presents a new way to identify anomalies based on wavelet transform and identify outlier by the use of the wavelet transform modulus maxima , then pass the amendment of the outlier through inverse transform the wavelet transform coefficient. Evidence shows that this method can be used to identify and correct the two types of outlier simultaneously and the results are obvious.