通过压缩感知框架对大规模不规则采样地震数据进行重构：一种基于 Seislet 变换的 FPOCS 方法
Recompositing of Vast Irregularly-Sampled Seismic Data via Compressed Sensing Framework: An FPOCS Based on Seislet Transform Approach
AI总结：
本文提出在 seislet 变换域结合稀疏约束的 FPOCS 压缩感知方法，用于重构不规则采样地震数据，实现更快收敛、更少迭代和更优信号恢复。
摘要：
从不规则地形表面采集地震数据时，往往会受到不规则且非等价的震源—检波器排列的制约，甚至在存储原始信号后产生坏道。在地震数据预处理方面，我们必须从给定信号中提取尽可能多的信息，从而使后续处理和解释能够获得极其准确的结果。我们将压缩感知定理应用于 Sigmoid 大规模不规则采样地震数据，所采用的方法是在 seislet 变换域中具有稀疏约束的快速凸集投影（fast projection onto convex sets，FPOCS）算法；与其他常规方法相比，该算法收敛更快，并能保持最佳的信号恢复效果。正如所证明的，由于 seislet 变换域中存在稀疏得多的结构，FPOCS seislet 变换方法相比其他方法能够取得准确且更高质量的数据恢复结果。此外，FPOCS 算法还能有效减少达到最佳数据填补所需的迭代次数。
英文摘要：
Acquiring seismic data from irregular topographic surface is oftently oppressed by irregular and nonequivalent source-receiver arrays and even more it yields bad traces after storing the original signal. In the light of preprocessing seismic data, we have to extract out most of the given signal, thus further processing and interpretation can obtain extremely accurate outcomes. We applied Compressed Sensing theorem on Sigmoid vast irregularly-sampled seismic data based on the fast projection onto convex sets (FPOCS) algorithm with sparsity constraint in the seislet transform domain, which gives faster convergence than other conventional methods and is preserving an optimum signals recovery. The FPOCS seislet transform approach can achieve accurate and high data recovery results than other methods because of a much sparser structure in the seislet transform domain as demonstrated. Moreover, FPOCS algorithm is also efficient in minimizing the number of required iterations to achieve optimum data refilling.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
