Abstract:During the propagation of seismic waves underground, the high-frequency seismic response of thin reservoir is absorbed and attenuated, which poses a challenge in seismic thin reservoir prediction. The high-resolution processing techniques have the capability to significantly expand the frequency range of the seismic data, so it becomes a key technique for thin reservoir prediction. Most of these techniques necessitate the extraction of seismic wavelets. However, the spatial and temporal variations of seismic data result in multiple solutions for wavelet extraction. Simultaneously, the majority of techniques fail to consider the influence of spatial tectonic features on the high-resolution processing. In this paper, we propose a novel solution to address these two fundamental challenges by utilizing seismic spectral expansion, sparse reflection coefficients, and spatial continuity constraints. First, we propose an innovative spectral fitting method that aims to expand the frequency bandwidth while adhering to the desired wavelet constraints. This method allows us to fully utilize the effective frequency information. It not only obtains broadband seismic data but also captures precise wavelets. Then, sparse deconvolution is employed to further extend the frequency range by utilizing the accurately expected wavelet and obtaining a high-resolution reflection coefficient. Finally, the Hessian matrix regularization is employed to constrain the spatial continuity of the reflection coefficient. This method is validated in both the model and real seismic data. Compared to traditional sparse deconvolution and spectral modeling deconvolution with spatial constraints, this method not only expands the frequency bandwidth and enhances seismic resolution but also preserves operational frequency information and improves the spatial continuity of seismic data. It has been verified that this approach can be used to forecast thin reservoir and reconstruct spatial tectonic characteristics.

Blind Sparse-Spike Deconvolution with Thin Layers and Structure

Multi-frame Blind Deconvolution Using Sparse Priors

Sparse Deconvolution Method for Ultrasound Images Based on Automatic Estimation of Reference Signals.

Blind Deconvolution Using an Improved L0 Sparse Representation

High-precision Seismic Data Reconstruction with Multi-Domain Sparsity Constraints Based on Curvelet and High-Resolution Radon Transforms

Guaranteed Blind Sparse Spikes Deconvolution via Lifting and Convex Optimization

CSwT-SR: Conv-Swin Transformer for Blind Remote Sensing Image Super-Resolution with Amplitude-Phase Learning and Structural Detail Alternating Learning

Single Image Blind Deconvolution Using Sparse Representation and Structural Self-similarity

Sparse Multichannel Blind Deconvolution of Seismic Data via Spectral Projected-Gradient

A Sparse Spike Deconvolution Algorithm Based on a Recurrent Neural Network and the Iterative Shrinkage-Thresholding Algorithm

Sparsity-Promoted Blind Deconvolution of Ground-Penetrating Radar (GPR) Data

Borehole-driven sparse-spike deconvolution and seismic bandwidth enhancement

Sparse Blind Deconvolution with Nonconvex Optimization for Ultrasonic NDT Application

Multichannel deconvolution with a high-frequency structural regularization

Blind deconvolution subject to sparse representation for fluorescence microscopy

Data and Model Dual-Driven Seismic Deconvolution via Error-Constrained Joint Sparse Representation

Learning the Blending Spikes Using Sparse Dictionaries

Blind spectral inversion of seismic data

A novel high-resolution imaging method based on sparsity in the time domain and spectral fitting in the frequency domain

Generalized MED blind deconvolution of GPR data and its sparsity-promoted solution

Improving Time Estimation by Blind Deconvolution: with Applications to TOFD and Backscatter Sizing