三维叠后零相位同态反褶积

Zero-phase homomorphic deconvolution after 3-D stack

在提高分辨率的处理中，人们比较熟悉叠前统计子波反褶积和地表一致性反褶积，它们在提取地表一致性信息和提高分辨率方面起到了十分有效的作用。然而在叠后处理中往往采用单道反褶积方法，但是，从高保真角度讲，采用二维和三维叠后多边或一体化反褶积将更加必要。本文首先应用自组织神经元聚类和分形果类方法求出对应地层组的三维一体化子波，然后完成叠后三维零相位固态反褶积处理，从而达到保持地质异常作信息、提高分辨率的目的。

In high resolution processing, people often use statisticat wavelet deconvolution before stack and surface-consistent deconvolution, which play very effective roles in extracting surface-consistent informations and improving seismic data resolution. However, single-trace deconvolution is usually used in post stack data processing. In view of angular high fidelity, multi-trace or all-trace deconvolution after 2-D or 3-D stack will be very necessary. First the 3-D all wavelets relating to strata are obtained by making self-organized neurone clustering and fractal clustering,then the zero-phase homomorphic deconvolution after 3-D stack can be achievedsatisfactorily. As a result, the informations of abnormal geology body can be retamed, and seismic resolution improved successfully.