摘要
本文从时频分析的角度出发，通过分析地震波与高频噪声在地震记录上的空间分布特征、频率分布征和能量分布特征等方面的差异，利用统计分析方法，提出了一种自适应的高频噪声的检测准则和压制方法。即先统计求取地震记录的正常子波振幅谱，然后据此识别出含有高频噪声的异常振幅谱，再对其进行压制。实际应用效果表明，该方法能有效地压制各类高频噪声，并保真地震记录的其它特征；同时，还具有适应性强、效果稳健等特点。
Having known the distribution,frequency and energy differences between effective seismic wave and high frequency noise by making time-frequency analysis,we advance a statistics-based new method for self-adaptive detecting and suppressing high-frequency noises. In the method,the amplitude spectra of normal wavelets in seismic data are produced statistically, then the abnormal amplitude spectra where high frequency noises exist can be found and suppressed desirably. It is shown that besides good adaptability and stability, the method effectively suppresses various high frequency noises and offers satisfactory fidelity in other characters of seismic data.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
1997年第3期318-326,共9页
Oil Geophysical Prospecting
关键词
高频噪声
时频分析
地震资料
自动识别
high frequency noise,amplitude spectrum, time-frequency analysis,statistics, recognition, suppression
