SEISMIC NOISE ATTENUATION BY TIME-FREQUENCY PEAK FILTERING BASED ON BORN-JORDAN DISTRIBUTION

The signal-to-noise ratio (SNR) improvement of seismic signals is an important step in seismic exploration. A time-frequency peak filtering based on Born-Jordan distribution (BJD-TFPF) algorithm is proposed in this paper. Time-frequency peak filtering (TFPF) is an effective filtering method for signal enhancement. The Born-Jordan distribution (BJD) has a good balance between suppressing cross terms and maintaining time-frequency aggregation. This algorithm takes advantage of TFPF and BJD to reduce noise while preserving the signal. In this paper, we also discuss different factors affecting the BJD-TFPF in terms of improving the practicality. The proposed method is applied to synthetic records and real seismic data. The results show that this algorithm can recover the signal effectively and is suitable for the filtering of seismic data. The results also reveal the effects of various factors on the practicality of the BJD-TFPF.