Research On Mechanism And Application Of Mathematical Morphology In Suppressing Coherent Noise In Seismic Exploration

Seismic exploration is the most important,widely used,and most effective method in geophysical exploration to solve oil and gas exploration problems.Its workflow can be divided into three parts: seismic data acquisition,seismic data processing and seismic data interpretation.Seismic data processing is an important part of seismic exploration.The quality of seismic data processing results will directly affect whether subsequent work such as seismic data interpretation can be carried out smoothly.Subject to the influence of the acquisition environment,the seismic data collected in the field often contains a lot of noise.Therefore,whether the noise in the seismic data can be removed has become a critical step in seismic processing.The research of noise suppression methods has always been a hot topic in geophysical prospecting.Denoising methods based on signal characteristics,mathematical transformations,sparse expression and wave theory have all been relatively mature applications.In recent years,mathematical morphological filtering(MMF)has been introduced as a new filtering method into seismic signal processing.It has quickly become a research hotspot due to its good time-space domain waveform separation advantages.For both random noise and coherent noise in seismic data,mathematical morphology filtering has a good signal-to-noise separation effect.This article will make some targeted research on the existing mathematical morphological filtering technologies as follows:Starting from the algorithm principles and basic operations of mathematical morphology,systematically summarized mathematical morphology filtering and a series of related methods and techniques,and combined its operation theory with seismic exploration,and the correlation between mathematical morphological filtering,structural element and seismic signal is exploratoryly proposed.Through the analysis of noise,the mechanism in the filtering process of seismic exploration is sorted out in detail.In-depth research and application promotion are carried out for the mathematical morphology filtering algorithm.Traditional mathematical morphological filtering can often only filter for the time continuity or spatial continuity of the signal.This article expands it to high-dimensional mathematical morphological filtering based on the operational meaning and filtering mechanism of traditional mathematical morphological filtering.The basic algorithm and operational meaning of high-dimensional mathematical morphological filtering are described.Through the expansion of the operational dimension,the application field of high-dimensional mathematical morphology filtering is extended to high-dimensional seismic data,and applied to the filtering process of white noise,band-limited random noise and abnormal amplitude values.The article verifies the superiority of mathematical morphological filtering by comparing the denoising results with traditional methods.At the same time,for the complex types of coherent noise in seismic exploration,combining the kinematic mechanism of noise,time-space domain characteristics and mathematical morphological filtering for coherent noise.A trajectory based spatial mathematical morphological filtering(TS-MMF)technique is proposed.This method is suitable for the signal-to-noise separation of complex signals.The reliability of the method is verified by the application of the signal-to-noise separation of marine surface diffraction noise and internal multiples in seismic data.Based on the characteristic analysis of the coherent signal in the frequency domain,and through the study of the mechanism of traditional mathematical morphological filtering,the application of mathematical morphology filtering in the frequency domain is innovatively proposed.Using the different phase change laws between the effective reflected signal and noise in the frequency-space domain,a method for suppressing random noise and coherent signal reconstruction by frequency domain mathematical morphology is proposed,and the feasibility of the method is demonstrated through practical data application examples.