Seismic weak signal extraction based on blind signal processing

Blind signal processing technique is one of the hot topics in the field of modern signal processing, aiming at solving problems such as how to separate or estimate the waveforms of the original source from an array of sensors or transducers without or with little knowledge of original waveforms and the characteristics of transmission channels. This paper presents the application of blind signal processing technology to the extraction of seismic weak signals. Based on the investigation and analysis of the relationship between blind signal processing theory and seismic reflection features of subtle traps, an aliasing model of seismic blind-source signals is established in order to extract weak signals from seismic data of target reservoirs. At same time, two new strategies for weak signal extraction are proposed. Using reflection similarity among seismic traces of surrounding rocks and reflection differences of weak signals from target reservoirs, we developed an iterative algorithm for weak signals extraction. The results of simulation and seismic data processing show that the method can extract seismic weak signals successfully and thus improves the resolution of seismic data.