基于结构匹配反褶积的同时震源数据直接成像

针对同时震源数据最小二乘偏移(LSM)计算效率低的问题,笔者研究并提出了同时震源混合数据直接成像的结构匹配反褶积(SMD)方法。通过分窗的方式求取局部Hessian逆,加入局部结构约束,可以压制混合数据成像串扰噪音的同时对成像做照明补偿,提高成像质量。并在复杂sigesbee2A模型的数值测试中验证了SMD方法的有效性。结果表明,SMD能够有效地压制成像中的串扰噪音,且对成像振幅有一定的补偿作用,计算效率相比LSM有巨大提升。

Simultaneous-source data direct imaging by structure-matching deconvolution

A structure matching deconvolution (SMD) method for direct imaging of simultaneous-source blendeddata is studied and proposed to solve the problem of low computational efficiency of simultaneous-source data least square migration (LSM).The local inverse Hessian is obtained by dividing the image into windows, and the local structural constraint is added to the LSM process. The method can suppress the blended data imaging crosstalk noise, compensate the illumination and improve the imaging quality at the same time.The validity of the SMD method was verified in the numerical tests of the complex sigesbee2A model.The results show that SMD can effectively suppress the crosstalk noise and compensate the imaging amplitude to certain extent. Moreover, the computational efficiency is greatly improved compared to LSM.