Full text loading...
-
Micro-seismic Data Denoising Based on Sparse Representations over Learned Dictionaries
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 77th EAGE Conference and Exhibition 2015, Jun 2015, Volume 2015, p.1 - 5
Abstract
The ground microseismic monitoring has characteristics of less microseismic events and lower signal-to-noise ratio. So it is necessary to suppress the random noise and improve the SNR if we want to enhance the accuracy of first break picking and the reliability of inversion results. In this paper, we study a denoised method in microseismic data based on sparse representation over learned dictionaries. Compared with traditional Fourier transform, wavelet transform and curvelet transform using some certain basis functions, it gets a data-based dictionary according to the input data by training the initialized dictionary. Then the denoising can merge into the dictionary learning naturally. We apply the method to a synthetic and real data. The denoised results show that it is effective to the noise attenuation without changing the first arrival time in microseismic data. The SNR has also improved significantly.