Application Of Wavelet Transform In Removing Random Noise Of Seismic Data

Improving signal-to-noise ratio(SNR) is an important step in seismic data processing. To achieve this, it is necessary to remove noises from seismic data. Noises of seismic data mainly are made of coherent noises and random noises. Generally speaking,appearance of coherent noises have is regular , particular methods can be used to remove them according to their features. However, random noises have no features ,so it is difficult to remove them when compared with coherent noises. The main work of this paper is reducing random noises in seismic data by analyzing characters of seismic data and using wavelet transform method to achieve the purpose of improving SNR of seismic date.Wavelet transform is a excellent mathematical analysis tool which developed in the late eighties of the last century, Because of its good characterization of time-frequency localization of the signal, it can be carried out on any frequency components fine description of a"mathmatical microscope,"said.In recent years, it has veen widely used in many areas such as signal de-noising .Wavelet transform is the center of this article, based on this center, significance of the issues is described,meanwhile, basic theory of wavelet is elaborated.Continuouswavelet transform,discrete wavelet transform and multiple analysis are introduced . When the details of method of reducing noises are considered, we selected wavelet threshold denoising and spatially-selective-noise-filtration(SSNF) methods as studing object. And in wavelet thresholding denoising method,some key problems are discussed in detail, ??while, we proposed a new thresholding function..In the spatially-selective-noise-filtration(SSNF), in order to highlight the edge of a signal, advanced algorithm is introduced. Finally, simulation experiments and the test of actual seismic data indicated that the two methods of seismic data denoising can achieve agood results.