Active sampling design for seismic data based on dynamic matching distance

Seismic data acquisition and reconstruction based on compressed sensing can enhance exploration efficiency and reduce costs. The acquisition of seismic data is foundational for reconstruction, where acquisition can be viewed as a sampling operator, directly impacting the quality of the reconstruction results. To obtain a more optimal sampling operator, this study proposes an active sampling design for seismic data based on dynamic matching distance. This method involves deploying sampling points twice to complete the sampling design, with the data from the initial sampling guiding the placement of the remaining points, ensuring more sampling points are allocated in areas with complex structures. The efficacy of this method is validated through comparative tests of sampling and reconstruction of two-dimensional and three-dimensional seismic data against jittered sampling.