Research On Construction Method Of Relative Geological Time Based On Phase Estimation

In seismic exploration,interpretation of seismic horizon is the basic work of seismic interpretation,and the recognition and tracking of seismic horizon is a very important link of seismic horizon interpretation.Relative geologic time(RGT),as a way to identify and track seismic horizons,is very important for seismic horizon tracking and sedimentological interpretation.The traditional method of generating RGT is to manually identify as many horizons as possible,and then interpolate the identified horizons to obtain this method.Although this method is direct,the accuracy is not high,and it relies heavily on subjective experience.In order to solve the above problems,a RGT construction technology with better performance was developed,and the phase estimation method was applied to generate RGT,and good results were obtained.Traditional phase estimation method is usually divided into two steps: first,from the observed image through a instantaneous phase noise surrounding the operation phase without noise image package,then the first step to get package phase unwrapping phase to estimate the absolute phase of the image Compared to the manual interpretation,phase estimation method has better ability of discontinuous maintain and higher computational efficiency.In this paper,by studying the phase unwrapping technology and the corresponding relationship between RGT and horizon in seismic data,the RGT construction method based on phase estimation is designed.Specifically,the main innovative work of this paper is as follows:(1)In order to solve the problem that the data quality affects the accuracy of RGT construction,in this paper,the block matched filtering(BMF)algorithm and tensor singular value decomposition(T-SVD)are fused to improve the first step in the phase unwrapping process,and then the graph cut phase unwrapping max-flow(PUMF)algorithm is combined to estimate the absolute phase of the image.A tensor block matched filtering graph cut phase unwrapping max-flow(TBMF-PUMF)algorithm based on tensor block matched filtering is proposed to construct RGT for seismic data.The algorithm focuses on extracting the enveloping phase of noiseless image from the instantaneous phase surrounded by noise in the observed image.By transforming seismic data into complex data,a complex domain suitable for sparse representation of complex data is constructed by using T-SVD decomposition,and a data adaptive data processing method of orthonormal complex domain is designed.The algorithm includes five steps: image block grouping,T-SVD decomposition,spectrum hard thresholding,image block reconstruction and image block aggregation.The PUMF algorithm is applied to the wrapping phase after denoising.For each precision,we map the problem into a binary optimization sequence,construct the appropriate graph by graph cutting method and calculate the minimum energy to select the best cutting path.Therefore,by using the unified principle of phase unwinding and denoising applied to seismic data,the generated RGT can achieve strong noise attenuation while preserving image details.(2)For pixel connection influence accuracy of structuring the RGT,PUMF algorithm is presented in this paper,on the basis of improving the second step in the process of phase unwrapping,by adding a constraint horizon of way of pixels in the process of phase unwrapping connection mode,based on the horizon constraints of tensor block matched filtering graph cut phase unwrapping max-flow(TBMF-HPUMF)algorithm.The algorithm focuses on unwrapping the wrapped phase of the noiseless image to estimate the real phase of the image.Based on the TBMF-PUMF algorithm,horizon constraints are added to improve the accuracy of absolute phase estimation,so as to improve the accuracy of RGT construction.By adding a horizon of manual interpretation to the wrapping phase unwrapping process,the connection mode of data points in the whole phase unwrapping process can be changed.This method not only inherits the fast performance of the minimum cutting process of graph cutting method,but also effectively optimizes the PUMF algorithm,so as to obtain a more geological RGT.