改进阈值的Curvelet变换稀疏反褶积

Sparse deconvolution based on Curvelet transform of improved threshold

依据反射系数在曲波域更为稀疏的特点,在曲波域实现稀疏反褶积,同时利用地震记录中反射系数在空间方向的相似性约束反演结果。对数据中的噪声采用一种改进的阈值函数,具有较好的光滑性,通过参数的合理调节,使阈值函数在硬阈值和软阈值之间灵活变化,将改进阈值函数应用于Curvelet变换稀疏反褶积。模型和实际资料处理结果表明:改进阈值的Curvelet变换反褶积方法能较好地反演反射系数,且具有一定的抗噪性,能较好地压缩地震子波,提高地震资料的分辨率。

Since reflection coefficients in the Curvelet domain are rather sparse,we can achieve sparse deconvolution in the Curvelet domain,and at the same time use the spatial similarity of reflection coefficients to constraint seismic data inversion results.An improved and smooth threshold function is used for denoising,and parameters can reasonably be adjusted to flexibly change the threshold function between the hard threshold and the soft threshold.The improved threshold function is applied for sparse deconvolution in the Curvelet domain.Processing tests on both synthetic and real data show that the proposed approach performs well in the reflection coefficient inversion,has a certain resistance to noise,and can well condense wavelet to enhance seismic data resolution.