Random noise decaying of seismic data based on steerable pyramid decomposition

摘要

In the paper, the basic principle of decomposition and construction of Steerable Pyramid, which has multi-scale and multi-direction characteristics, was introduced in detail. It was applied in the random noise decaying of seismic data based on adaptive BayesShrink threshold, and gone on emulation and real seismic data processing. In order to compare the effect of noise decaying, the effect of noise decaying of Steerable Pyramid decomposition based on adaptive BayesShrink threshold was compared with that of wavelet transform based on adaptive BayesShrink threshold. The result proved that Steerable Pyramid decomposition based on adaptive BayesShrink threshold could relatively completely remove the noise while the edge of image kept well and the detail part also kept at the same time. Other thresholds are researched now for better effect of noise decaying. The result of noise decaying was good and easy to realize, so that Steerable Pyramid decomposition has the feasibility and prospect in the deal with seismic data.