Random Noise Attenuation Using Variable Q-Factor Wavelet Transform
M. Irani Mehr and M.M. Abedi
Event name: 79th EAGE Conference and Exhibition 2017
Session: New Paths in Noise and Multiple Attenuation
Publication date: 12 June 2017
Info: Extended abstract, PDF ( 1.2Mb )
Price: € 20
Seismic data have variable characteristics in time and location. Overlooking this important characteristic will reduce the effectiveness of any signal processing tool. In this study we use a new strategy by use of Rational-Dilation Wavelet Transform (RDWT). This method has the capability to achieve variable frequency resolution, which makes it an effective tool in suppressing random noise. The characteristics of the proposed transform change with time and location. This transform is invertible and it provides different Q-factors (Q-factor is the ratio of wavelet’s central frequency to its bandwidth). We adapt RDWT in time and space so it uses diverse range of Q-factors in accord with dominant Q-factor of a trace. We applied it to synthetic data and high frequency shallow Sub-Bottom Profiler data to evaluate the effectiveness of the method. We find outcomes of RDWT with variable Q-factoring time and space to be in agreement with frequency content of the data which indicates that the method is able to improve the continuity of the events by reducing the random noise content of the data.