Research On Structure-preserving Filtering Based On Mathematical Morphology

Seismic data processing technique is a very important part in the research of seismic exploration methods,in which the study of noise suppression is indispensable,and it is very important to improve the quality of seismic data and enhance the accuracy of interpretation.This paper mainly discusses the application of mathematical morphological filtering method,adaptive filtering method and structure-preserving filtering method based on mathematical morphological filtering in noise suppression and structural protection.The basic theory of mathematical morphology in seismic data processing is emphatically studied.Based on the morphological scale attribute,it suppresses coherent noise,low frequency noise and random noise on pre-stack data,and suppresses random noise on post-stack data.Besides,the signal-to-noise ratio estimation method is studied,considering the edge and structure information,using edge-preserving smoothing method to improve it,and based on the improved local noise level combined with morphological filtering method to suppress random noise.In addition,the structure-preserving morphological filtering method for 3D post-stack data is studied.Two-dimensional morphological filtering combined with direction is used to suppress random noise.The minimum standard deviation is the similarity evaluation standard.The model and real data processing experiments show that the above-mentioned methods are effective and have a good effect of suppressing noise.At the same time,the effective signal is well protected,and the edge and structural information are retained.