基于字典学习算法的地震资料去噪处理应用研究

Research on application of seismic data denoising processing based on dictionary learning algorithm

针对地震资料的质量一直是制约数据处理解释结果的重要因素,而且随着新技术的开发,采集到的数据量也在不断加大,噪声的存在也不可避免,因此去噪成为了一大问题,这里应用压缩感知理论,分别以DCT与K-SVD学习字典为稀疏基对不同层位的模型数据进行测试,之后对某工区的实际地震剖面进行处理。结果表明:DCT字典去噪时会损害原地震信号的高频信息,而K-SVD学习字典的效果明显优于DCT字典,在去除噪声的同时对原地震信号的信息能够很好地保留,但由于迭代次数的增加,处理时间较长,应用基于压缩感知理论的OMP算法进行迭代运算,加快了K-SVD学习字典的计算速度,且能重建恢复原信号,为后面的解释工作打下良好的基础。

The quality of seismic data is always an important factor in restricting the results of data processing and interpretation.With the development of new technologies,the amount of collected data is also increasing,and the existence of noise is inevitable.Therefore,denoising became a major problem.To overcome this problem,the compressed sensing theory is applied in the article,and the DCT and K-SVD learning dictionaries are used as sparse bases to test the model data of different horizons.After that,the actual seismic profile of a certain work area is processed.The results show that the DCT dictionary can be used to denoise.Damage to the high-frequency information of the original seismic signal,and the K-SVD learning dictionary is significantly improved compared to the DCT dictionary.It can remove noise while retaining the information of the original seismic signal.However,due to the increase in the number of iterations,the processing time is increased.The application of OMP algorithm for iterative calculation speeds up the calculation speed of the K-SVD learning dictionary;and it can reconstruct and restore the original signal,laying a good foundation for the subsequent interpretation work.