地震信号的小波去噪方法研究-地球探测与信息技术专业论文.docxVIP

Study Study On methods ofnoise removing by wavelet and their applications to seismic data processing Abstract Improving the signal—．to．-noise radio is an important task in the seismic data processing，therefore the study of de—noising methods is always hot in seismic exploring．With the development of digital signals processing technology,many excellent de-noising methods come forth．In the seismic data de·noising．there is all 油_portant research aspect that choosing the suitable de—noising method based on the feature ofseismic data． Wavelet analysis becomes a powerful tool in the field of nonstationary signals processing in the last few years，because ofits good ability oftime-frequency analysis． In the sanle time，many de—noising methods based on wavelet transform are coming OUt．Furthermore they ale proved to be superior to traditional Fourier transform de—noising methods．Four wavelet transform de-noising methods are studied in this paper．The processing results indicate that the Mold maximum value wavelet transform de—noising has advantage in the low signal—to—noise ratio date，and the method fits in the condition that contains white noise and many singularities．Spatially Selective Noise Filtration is more fit in the of higher signal—to—noise ratio，and its capability of edge reconstruction is better；Besides the WaveShrink method has widespread applicability and its effect ofde-noising is also very good；the translation invariant wavelet threshold value de--noising method can restrain pseudo-·Gibbs form WaveShrink threshold de-noising method,and get better Visual effect． The paper applies the four methods upwards to the seismic data proeessing， proves the validity of these methods through synthetic seismic recording and the practical seismic data,then achieves the goal of improving the signal-to—noise radio in seismic data processing．． Key words：Seismic data,Random noise，De-noise，Wavelet transform n 原创性声明本人声明，所呈交的学位论文是本人在导师指导下进行的研究工 原创性声明 本人声明，所呈交的学位论文是本人在导师指导下进行的研究工 作及取得