Research On Seismic Signal Denoising Technology Based On Wavelet Transform

The acquisition of the seismic data mainly contain three parts: the collection,processing and interpretation.A lot of interference noise comes from the the detection layer depth,complex working environment,and the cause of the instrument itself.So how to eliminate the interference noise is the primary task of data processing,because a good SNR can guarantees the accuracy of the data interpretation.Compared with the traditional Fourier transform,wavelet overcomes the drawback of the single resolution analysis,and the signal features can be viewed on different scales.It also has a good ability of local analysis in time domain and frequency domain.Nearly 30 years,the wavelet transform theory has been developed very quickly.In this paper,the main work is to research the related theory of wavelet transform and its applications in seismic signal denoising.In this paper,we first illustrates the significance of seismic signal denoising and introduce the research status of it.And then we introduce the two main noise of the earthquake: coherent noise and random noise,as well as the causes of different noise,their respective characteristics and the commonly used method to remove the noise? And finally we leads to the focus of this paper about wavelet transform denoising.And then the wavelet analysis theory is introduced in detail,including compared with Fourier transform and short-time Fourier transform,definition of wavelet transformation,the principle of continuous wavelet transform and discrete wavelet transform,the concept of multi-resolution analysis,and the implementation of the classic Mallat algorithm,and so on.The further research is about the application of wavelet transform in seismic signal denoising:Firstly,through the error analysis,the bior2.4 wavelet whose decomposition and reconstruction mother wavelet are different is selected,and the scale of wavelet decomposition is determined.Then we introduce two methods of wavelet transform: modulus maxima denoising and threshold denoising.And with the aid of Matlab platform,we completed the simulation of different denoising method.Finally,we use the real seismic signal to verify it.And we analyze the data and processing effect figure,Proved that the effectiveness of the wavelet transform in seismic signal denoising and points out the still existing problems.