自适应信号重构技术提高地震数据的信噪比

The Auto-adaption Signal Reconstruction Technique to Improve the Signal to Noise Ration of Seismic Data

利用信号和噪声的特征值差别,采用奇异值分解技术压制噪声,提出了基于多道相似系数大小来自适应地选择特征值个数,进而进行信号重构,分离信号和噪声,从而达到提高信噪比的目的。通过对合成数据和实际地震数据的处理结果对比表明该方法能最大程度地保持信号,从而有效地提高地震资料的信噪比。

The differences of eigenvalue between signal and noise are used in this paper. The noise is suppressed by singular value decomposition technology. The method is put forward that the eigenvalue number is auto--adaptively chosen based on multi--channel semblance coefficient size. The signal is reconstructed and the signal and the noise are separated, so the signal--to--noise ratio is improved. The results of synthetic data with noise and practi- cal seismic data processing show that this method can further hold the signal and effectively improve the signal to noise ration of the seismic data