Abstract: The random noise is a kind of noise with wide frequency band in seismic data and influences the useful signal very much. It is very important to remove the random noise from the seismic data. Fast Fourier Transform (FFT) is the traditional method to remove the random noise and can analysis a signal on the whole spectrum, however, cannot analysis nonstationary signal. Wavelet analysis is a novel method to improve the practical data and remove the noise and is prior to other methods. This paper uses the wavelet analysis to remove the random noise from the seismic data, the results show that wavelet analysis can more efficiently remove the noise and improve the precision of seismic data.

Key words: random noise, wavelet analysis, Wiener filter, FFT, denoise