Compressive Sensing Reconstruction Of Seismic Data Based On Curvelet

In recent years, with the continuous development of oil exploration, the structure and environment of the exploration target area are becoming more and more complicated, which exacerbates the irregular and incomplete situation of the seismic survey data,influences the processing and interpretation of the data, and finally affects the oil and gas judgment. In the past, the traditional reconstruction method adopted by the Nyquist sampling theorem requires a high sampling rate, and in the face of complex exploration conditions can not do the appropriate collection and adjustment, the previous exploration costs are very large. For this phenomenon, we need to develop a good reconstruction algorithm, as far as possible to achieve a complete seismic image reconstruction, to make the utilization of seismic data can be improved. In this paper, Curvelet transform and compression perception are combined with each other, and the related algorithm is proposed to reconstruct the seismic image and enhance the visual quality. The main research contents are as follows:1.Seismic Image Reconstruction Based on Curvelet Shrinkage Threshold Iteration Algorithm. Combining the theory of compression perception, wavelet transform,DFT transform and Curvelet transform are analyzed, and the seismic image is reconstructed.Curvelet transform has a good multi-scale geometric analysis ability, it can have the edge of the curve of the seismic image of the optimal sparse expression. A seismic image area is divided into multiple sub-regions, to achieve a certain interval of random sampling. Finally,an adaptive bivariate shrinkage threshold iterative reconstruction algorithm based on Curvelet transform is designed according to the characteristics of Curvelet transforming high frequency subband information entropy. The results show that the proposed algorithm has good reconstruction effect in seismic images.2.Research on Seismic Image Reconstruction Based on Bregman Iterative Algorithm.The Bregman iterative algorithm, the Bregman iterative algorithm, and the residual Bregman iterative algorithm are used to compare the advantages and disadvantages of the Bregman iterative algorithm, the Bregman iterative algorithm, and the Bregman iterative algorithm.,And an improved Bregman iterative algorithm is proposed. In the Bregman iterative framework, the soft threshold is used as the threshold operator H, and the threshold parameter selection based on the H-curve criterion is proposed to improve the accuracy of seismic image reconstruction. Compared with the experimental results, the improved Bregman iterative algorithm has a good reconstruction effect in the seismic image.3. Research on Compressed Sensing Observation Matrix. According to the characteristics of seismic images, several commonly used observation matrices are selected to analyze the characteristics of these five observation matrices, and the generalized rotation matrix is the main object of discussion. Under the principle of its construction, the generalized rotation matrix has a strong stability, and its performance is better than other observation matrices. In addition, the generalized rotation matrix is also a deterministic observation matrix, which is easy to realize and store. It is found that it has a strong column non-correlation and enhanced low-band sampling. That is, in order to enhance the non-correlation between columns and columns can modify the observation matrix matrix of the first half of each part of the element coefficient, while modifying the coefficient value, enhanced the low frequency band sampling.Finally, the experiment of seismic image reconstruction is carried out. The experimental results are compared and the reconstruction effect of the generalized rotation matrix is the best.