二进小波变换方法的地震信号分时分频去噪处理

SEISMIC DATA TIME-FREQUENCY DOMAIN DENOISING BASED ON DYADIC WAVELET TRANSFORM

应用二进小波变换理论，对地震记录剖面进行小波分解，在小波变换域，对各分量剖面进行分时分频相关分析，并依此进行加权处理，最后通过小波反交换算法恢复出地震记录剖面．经实验该方法对衰减地震信号的噪声具有比较好的效果．

A new method for seismic data denoising isproposed in this paper. We decomgose every trace of a seitmic section using dyadic wavelet transform, so several sub-sections are formed in time-frequency domain. To each sub-section a correlation analysis is made and every trace of each sub-section is weighted according to the analytic result. It is calied weighting denoising processing.From the sub-sections we recover the seismic section in which noise has been compressed. We also illustrate the performance of the method using real-world seismic data.