Enhancing the resolution of seismic data based on the non‐local similarity

Abstract

ABSTRACTWe propose a resolution‐enhancement method based on the non‐local similarity of the seismic profile. Because of the similarity in underground structures, the seismic data have similar event structures in different regions. We introduce the similarity as prior information to enhance the resolution of seismic data. The similarity, as a regularization term for enhancing the resolution, can improve the stability of the solution. The non‐local similarity constraint is usually implemented based on nuclear norm minimization. We use the L1–L2 norm instead of the L1 norm in the nuclear norm, which will obtain a more accurate approximation of low rank. Compared with the deconvolution algorithm based on the L2 norm, we use non‐local similarity as a constraint term. The proposed method can enhance the resolution with a low signal‐to‐noise ratio, suppress random noise and improve lateral continuity. The synthetic and field data prove the superiority of the proposed method for low signal‐to‐noise ratio data.