Seismic Signal Filtering based on Pseudo Wigner-Ville Distribution and Catte Model

In practice, there are various noises included in the collected seismic data. As a common background noise, random noise may interfere even annihilate valid signals. In view of this, it needs to adopt effective methods to eliminate random noise as much as possible. This paper proposes a filtering method that is based on pseudo Wigner-Ville distribution (PWVD) and Catte model. PWVD is a windowed form of Wigner-Ville distribution (WVD) which can make the nonlinear signal to be locally linearized in addition to weakening the influence of cross item. And the Catte model is a kind of anisotropic diffusion algorithm which can eliminate the noise while protecting the edge of image. The implementation of the proposed method is to do PWVD for the frequency modulation form of noise-containing signal and then to filter the PWVD via Catte model, at last to adopt peak search to obtain the estimation of the valid signal.