文章摘要

文章信息/Info

Title:: Combining MFT and Curvelet Transform Method to Extract Weak Signal in Active Source of Air Gun

作者:: 谭俊卿¹; 杨润海^2; 向涯³; 王彬²; 姜金钟²; (1.云南大学地球科学学院,云南昆明 650091; 2.云南省地震局,云南昆明 650224; 3.中国地震局地震研究所中国地震局地震大地测量重点实验室,湖北武汉 430071)

Author(s):: TAN Junqing¹; YANG Runhai²; XIANG Ya³; WANG Bin²; JIANG Jinzhong²; (1.School of Earth Sciences,Yunnan University,Kunming 650091,Yunnan,China)(2.Yunnan Earthquake Agency,Kunming 650224,Yunnan,China)(3.Key Laboratory of Earthquake Geodesy,Institute of Seismology,China Earthquake Administration,Wuhan 430071,Hubei,China)

Keywords:: weak air gun signal of active source; Matched Filtering Technology; Curvelet Transform denoising; velocity changes

摘要:: 提出由一维模板匹配滤波技术(MFT)和二维曲波(Curvelet)变换法穿插的数据处理方法,即先通过一维模板匹配滤波方法得到相关系数,将相关系数组成二维数据并用曲波变换法处理,最后将各道对应相关系数分别与模板信号褶积,得到高信噪比恢复信号。将该方法用于处理仿真数据和宾川主动源气枪信号,结果表明:在宾川主动源气枪信号的处理中,本方法较单一数据去噪方法恢复能力更好,可对气枪主动源的模拟与实际信号进行更好的数据提取与恢复,得到噪声干扰更少的地下介质波速变化,提高低信噪比数据的利用率与可分析性。

Abstract:: In this paper,we introduced a data processing method interspersed with two method that based on one-dimensional template Matching Filtering Technology(MFT)and two-dimensional Curvelet Transform method.Firstly,the correlation coefficients are obtained by the one-dimensional MFT,then the correlation coefficients are composed into two-dimensional data and processed by Curvelet Transform method.Finally,the corresponding correlation coefficients are respectively folded with the template signal to obtain the recovery signal.Then,we applied this method to the processing of simulation signal and air gun signal of Binchuan active source.Experiments show that this method is more better than that processing by one single method,with better recovery ability in Binchuan Air Gun Experimental Base,which can get better recovery signal in simulation and actual recorded data signal processing,and get the wave velocity variation with less noise interference in underground media.It is beneficial to the analysis and research of the following seismologists and improve the utilization and analyzability of low SNR data.