Scholarlink

<p>Strong seismic acquisition footprints often hide subtle subsurface details and reduce the reliability of amplitude-sensitive attributes. Standard global filters face challenges in preserving structural sharpness, while the intensive computational demands of low-rank decomposition methods can limit their efficiency for large-scale applications. To address these issues, we propose the Robust-Reference Side-Window Filtering (RR-SWF) framework. The workflow begins by using an anti-leakage Fourier spectrum combined with an <i>L</i><sub><i>p</i></sub> norm to accurately estimate local structural dips, which are then used to construct a structural flattening domain. This step aligns variable subsurface layers horizontally, transforming the signal separation task into a straightforward morphological distinction problem. In this domain, we introduce a structure-guided pilot signal to govern the side-window selection. This reference mechanism prevents the filter from locking onto high-amplitude footprint noise, allowing us to decouple genuine horizontal reflections from steep footprint noise. We validated this strategy on synthetic models and complex 3D field datasets. The results confirm that the method increases the signal-to-noise ratio by eliminating intense footprint energy. It achieves this with negligible signal leakage, preserving the sharpness of faults. Furthermore, the algorithm is highly efficient. Its computational cost remains independent of the filtering kernel radius. The proposed approach is faster than iterative rank-reduction methods. This efficiency makes it a practical, high-fidelity solution for pre-conditioning massive 3D seismic volumes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Acquisition footprint suppression via robust-reference side-window filtering

  • 期刊论文

  • 发表日期：

    2026年2月21日

查看全文
  • Liang Huang,
  • Wenchao Su,
  • Yixin Xu,
  • Mingran Liu,
  • Jiangyun Zhang,
  • Xuhui Zhou,
  • Guoxu Shu,
  • Shoudong Huo

摘要

Strong seismic acquisition footprints often hide subtle subsurface details and reduce the reliability of amplitude-sensitive attributes. Standard global filters face challenges in preserving structural sharpness, while the intensive computational demands of low-rank decomposition methods can limit their efficiency for large-scale applications. To address these issues, we propose the Robust-Reference Side-Window Filtering (RR-SWF) framework. The workflow begins by using an anti-leakage Fourier spectrum combined with an Lp norm to accurately estimate local structural dips, which are then used to construct a structural flattening domain. This step aligns variable subsurface layers horizontally, transforming the signal separation task into a straightforward morphological distinction problem. In this domain, we introduce a structure-guided pilot signal to govern the side-window selection. This reference mechanism prevents the filter from locking onto high-amplitude footprint noise, allowing us to decouple genuine horizontal reflections from steep footprint noise. We validated this strategy on synthetic models and complex 3D field datasets. The results confirm that the method increases the signal-to-noise ratio by eliminating intense footprint energy. It achieves this with negligible signal leakage, preserving the sharpness of faults. Furthermore, the algorithm is highly efficient. Its computational cost remains independent of the filtering kernel radius. The proposed approach is faster than iterative rank-reduction methods. This efficiency makes it a practical, high-fidelity solution for pre-conditioning massive 3D seismic volumes.