The Application Of TFPF Based On Directionally Smoothed Pseudo Wigner Ville Distribution In Seismic Signal Processing

In practice, signals are often corrupted by noise, and this has the effect ofhindering the recovery of important information encoded in the signal. As seismicdata is complex and useful signal is under a good deal of noises, it’s hard to recoverthe original signal without loss. People have raise many methods to improve the SNR.Many signal processing algorithms work well for high SNR situations, but mostperform poorly when SNR decreases below a given threshold. So the appropriatealgorithm to remove random noise has become the primary task in seismicexploration.Time-frequency peak filtering (TFPF) is an effective algorithm in random noiseattenuation. In recent years, it has been developed and applied to seismic datadenoising. TFPF is a signal enhancement algorithm, which is based on the principle ofinstantaneous frequency estimation. TFPF is the technique that the noisy signal isencoded as the IF of an analytic signal by frequency modulated. Then, the signal isrecovered by taking the peak of time-frequency distribution (TFD) of analytic signal.For the nonlinearity of the seismic data, pseudo Wigner-Ville distribution (PWVD) isutilized to execute TFPF, which makes the IF in the window function be linear.Traditional time-frequency peak filtering used rectangular window to realize thelinearization. traditional time frequency peak filtering on time-frequency clusteringand noise suppression will be descend and will have yield a number of outliers. In thispaper, we construct a new family of TF distribution, namely, the joint distribution, toestimate the IF curves in order to reduce outliers in the cases of low SNR. Theconstruction of the joint distributions is based on the definition of the directionallysmoothed pseudo Wigner-Ville distribution (DSPWVD) and point wise adaptiveweight averaging of a bank of DSPWVD with different directions. The segments ofthe IF curve whose directions are close to that of the DSPWVD can be highlighted byeach DSPWVD and the entire IF curve will be enhanced by the joint TF distribution.In this paper，we use DSPWVD calculation time-frequency distribution, and then useTFPF instantaneous frequency estimation algorithm to estimate the signal, to removerandom noise, and keep the effective signal. To further improve the performance of the joint time-frequency distribution baseon DSPWVD. In this paper, using joint time-frequency distribution and PWVD oftime frequency peak filtering to process the seismic forward modeling and practicalseismic data respectively, and the processing effects of time frequency peak filteringfrom two kinds of distribution are compared, Including the time domain waveform,the earthquake wavelet frequency amplitude form one channel, the signal-to-noise ratio andso on. The results show that under the joint time-frequency distribution oftime-frequency peak filtering have higher estimation precision, it can suppress thestrong random noise, and at the same time it keeps the signal characteristics better,so it makes the signal-to-noise ratio improved further. It has great practical value forthe low SNR of seismic data processing.