文档介绍：
快速匹配追踪算法在地震信号去噪中的应
用#
余为维,冯磊**
5
10
15
20
25
30
35
(河南理工大学资源环境学院,焦作 454000)
摘要:地震信号的信噪比是决定地震资料品质优劣的关键因素,因此地震资料的去噪处理研
究也一直是研究热点。快速匹配追踪算法采用原子库分组式及树状结构搜索的方式分解信
号,相比传统匹配追踪,该算法分解快速高效。本文通过对该算法的研究,将其应用于地
震信号的去噪处理中,从合成记录及实际数据的处理结果来看,该技术有效地去除了噪声
干扰,达到了去噪目标。
关键词:信噪比;快速;匹配追踪;去噪
中图分类号:
The application of seismic signal de-noising technology
based on high-speed matching pursuit algorithm
Yu Weiwei, Feng Lei
(Resource and Environment School, Henan Polytechnic University, Jiaozuo 454000)
Abstract: The signal-to-noise ratio of seismic signal is the critical factor determining the quality of
seismic data, and therefore study on de-noising of seismic data has been a hot topic. High-speed
matching pursuit algorithm poses signals by adopting packet-mode atom dictionary and
tree-structured search. Compared with traditional matching pursuit, this algorithm is fast and
efficient. This paper studies this algorithm and applies it to de-noising processing of seismic signal,
the processed results of the synthetic record and the actual data indicate this technology removes
noise effectively and achieves the de-noising goal.
Key words: signal-to-noise ratio; high-speed; matching pursuit; de-noising
0 引言
在地震资料处理中高信噪比、高分辨率以及高保真度一直是其追求的理想目标,其中信
噪比又是高分辨率和高保真度的基础。提高信噪比在地震资料处理中显得尤为重要。目前的
常规去噪技术,对于具有一定时空分布规律的相干噪声可根据方向特性、相干性等予以去除,
如 f - k 滤波[1],对于分布无规律的随机噪声可采用如多项式拟合[2]等技