A modified Cadzow filtering algorithm based on anti-picket fence effect
Jianyou Chen, Wenchao Chen, Xiaokai Wang, Jiao Liu, Wei Wang · International Geophysical Conference, Qingdao, China, 17-20 April 2017 · 2017
Summary The Cadzow filtering algorithm has been widely used to suppress random noise from seismic data. This algorithm firstly transforms seismic data from time domain into frequency domain using DFT, use singular values decomposition (SVD) filtering in every constant-frequency slice, and inverse transform to time domain at last. But this algorithm suffers from picket fence effect in the DFT procedure, and it will weaken the denoising result quality at last. Therefore, this abstract proposes a modified Cadzow filtering algorithm based on anti-picket fence effect. The anti-picket fence effect is achieved by zero padding at the end of data and we confirm the zero padding length according to the denoising quality and algorithm computational efficiency synthetically. By using this modified Cadzow filtering algorithm based on anti-picket fence effect, we can get a better denoising result compared with the common method. This new method is tested on both synthetic and real datasets to show the efficiency.