A Study Of Radon Transform And Its Application In Blended Seismic Data Processing

Conventional seismic data acquisition process has long operation cycle and high exploration cost,while multi-source blended seismic acquisition technology can effectively reduce exploration cost.However,overlapping interference often occurs in blended seismic data after shot separation,and as with conventional data,data missing also occurs in blended seismic data.Radon transform is a common signal processing method,which is widely used in conventional seismic data processing,including denoising,multiple suppression and seismic data reconstruction.In this paper,we will study the application of Radon transform in blended seismic data and solve the problems in blended seismic data.In the field blended seismic exploration,due to the influence of terrain limitation,instrument failure and other factors,the regular sampling of data will be affected,and even the absence of data will occur.For seismic data processing,the lack of seismic data will cause serious impact,so it is very meaningful to study the reconstruction of mixed mining seismic data.In mixed mining seismic data,the processing of gun separation is often needed.The data after gun separation often have aliasing noise.It is very important to effectively eliminate aliasing interference for the result of gun separation.Radon transform plays an important role in de-noise and data reconstruction.It is very important to study the application of Radon transform in blended seismic data.Compared with the parabolic Radon transform,the hyperbolic Radon transform has higher accuracy.Since the in-phase axis of seismic data satisfies the assumption of hyperbolic curve,hyperbolic Radon transform can process the original seismic data directly,while parabolic Radon transform processes the seismic data after dynamic correction.Compared with conventional seismic data,the reconstruction of blended seismic data is more complex.First,the mixed seismic data should be separated,and then the separated single shot data should be reconstructed.Therefore,in the process of separation and reconstruction of blended data,high precision separation method and accurate reconstruction method are required.Therefore,in this paper,the median filter method is adopted in the shot separation of blended seismic data,and hyperbolic Radon transform is selected in the reconstruction of blended seismic data for processing.This paper first introduces the basic principle of Radon transform,and on the basis of discussing the conventional application scope of Radon transform,focuses on the application of Radon transform in blended seismic data,especially in the aspect of eliminating single-shot mixed noise and missing seismic trace reconstruction after the separation of mixed data.The amount of blended seismic data is relatively large,so the computational efficiency is particularly important.In the Radon transform processing,the operator solution using the conventional conjugate gradient method has low computational efficiency.In order to improve the solving speed,this paper introduces the FISTA algorithm in the operator solving.In the comparison of actual data processing,it is found that the FISTA algorithm is faster,significantly better than the conjugate gradient algorithm,and effectively improves the computational efficiency.In terms of improving the reconstruction accuracy,this paper introduces the sparse constraint method on the basis of hyperbolic Radon transform.Through the comparison with ordinary hyperbolic Radon transform,it is found that the sparse constraint can improve the resolution of data in the Radon domain and obtain more accurate reconstruction results.The results show that Radon transform can effectively eliminate the alias interference of blended data and reconstruct the missing seismic trace accurately.