Abstract
This paper deals with the noise reduction by the wavelet transform when the signal contain the noise of the impulsive and Gaussian components.In the case of the impulsive noise component, the algorithm of the noise reduction is the inverse finite impulsive response (IFIR). The estimate time of the IFIR is treated by the wavelet odd (or even) coefficients. On the other hand, the reduction of Gaussian noise component is carried out the treatment of low pass filter for the wavelet Fourier spectra. The effects of this method are investigated by the numerical calculation and are confirmed for the impulsive noise component. In addition, we apply the method for the seismic wave involved the impulsive noise and obtain the good results.
