反泄露傅里叶变换已在地震数据规则化、多维数据去噪及数据压缩等领域得到广泛应用,但其多在时间域实现,计算效率不高。对均匀网格反泄露傅里叶变换的常规实现方法进行了改进,在频率域给出一种均匀网格反泄露傅里叶变换的高效实现方法。该方法在频率域寻找傅里叶系数极大值及其位置,通过在频率域减去SINC函数,避免了常规实现方法在每次迭代过程中所使用的离散傅里叶变换及逆离散傅里叶变换,大幅减少了均匀网格反泄露傅里叶变换的运算量,提高了均匀网格反泄露傅里叶变换的计算效率。利用合成数据对方法进行了验证,结果表明,该方法不仅能够明显减少频谱泄露现象,提高稀疏表示性能,而且较常规实现方法提高了计算效率。
The antileakage Fourier transform (ALFT) has been widely used in seismic data regularization,multi-dimensional denoising,and data compression.However,the computational efficiency of regular grid ALFT is not high enough in the time domain to process massive seismic data,because it implements one discrete Fourier transform to convert the signal from the time domain to the frequency domain in each iteration.In this paper,we focus on an implementation of the ALFT in the frequency domain to improve its efficiency.The method searches for the maximum Fourier coefficient and subtracts the SINC function in the frequency domain,to avoid the discrete Fourier transform and the inverse discrete Fourier transform in each iteration.Tests on synthetic data show that the method can improve computational efficiency,while reducing discernible frequency leakage and improving sparse representation.
国家自然科学基金(41504092,41774135)、中国博士后科学基金(2016T90925,2015M572567)及中央高校基本科研业务费专项资金(xjj2016065)联合资助。