Research On Application Of Seismic Signal De-noising Based On Mallat Algorithm

There are many methods for seismic signal de-nosing, and those based on wavelettransform have a good performance and wider application. Moreover, there are vari-ous algorithms for implementing wavelet transform; Mallat algorithm is one of the mostcommon algorithms.In this paper, we focus on seismic signal de-noising. Because of unknown priorinformation of noise which a real seismic signal includes, in order to confirm that the de-noising method is effective, we simulate artificial seismic signals by using trigonometricseries and wavelet transform, and find that the artificial signals can re?ect some importantcharacteristics of the real seismic signals. Then we add random noise to the artificialsignals, study de-noising the noised signals and get a reliable de-noising method. At last,we de-noise a real seismic signal by the reliable de-noising method.We study the application of the de-noising by wavelet threshold based on Mallatalgorithm to process the seismic signals. According to the multi-resolution characteristicsof Mallat algorithm, we decompose the seismic signals using multi-scale wavelet Mallatdecomposition algorithm, process the wavelet decomposition coeffcients by different de-noising methods by threshold, reconstruct the signals using the reconstruction algorithm,and then get the de-noised signals. A lot of numerical experiments have been done onde-noising the artificial seismic signals using de-noising methods by wavelet thresholdhaving different parameters. By comparison with the corresponding results in relatedperformance parameters, we finally get a proper method to de-noise seismic signals, andthen we process the real seismic signal using the de-noising method. At last, we find thatthe processing effect is satisfactory.