DSU-Net: Dynamic Snake U-Net for 2-D Seismic First Break Picking

TL;DR

DSU-Net tackles the challenge of accurate 2-D seismic first-break picking by integrating Dynamic Snake Convolution into a U-Net framework to better capture horizontal continuity and FB jumps. The DSConv modules deform sampling along the horizontal or vertical axes, enabling robust extraction of tubular shallow features, and are most effective when placed in the first encoder layer. Evaluated on four public field datasets, DSU-Net achieves state-of-the-art accuracy and robustness, outperforming STA/LTA, a U-Net benchmark, and STU-Net in metrics such as HR@1px and RMSE, particularly under noisy conditions. Ablation studies confirm the efficacy of the dual-axis DSConv, the chosen extension scope and kernel size, and the proposed LMO-based data augmentation in improving generalization for FB picking.

Abstract

In seismic exploration, identifying the first break (FB) is a critical component in establishing subsurface velocity models. Various automatic picking techniques based on deep neural networks have been developed to expedite this procedure. The most popular class is using semantic segmentation networks to pick on a shot gather called 2-dimensional (2-D) picking. Generally, 2-D segmentation-based picking methods input an image of a shot gather, and output a binary segmentation map, in which the maximum of each column is the location of FB. However, current designed segmentation networks is difficult to ensure the horizontal continuity of the segmentation. Additionally, FB jumps also exist in some areas, and it is not easy for current networks to detect such jumps. Therefore, it is important to pick as much as possible and ensure horizontal continuity. To alleviate this problem, we propose a novel semantic segmentation network for the 2-D seismic FB picking task, where we introduce the dynamic snake convolution into U-Net and call the new segmentation network dynamic-snake U-Net (DSU-Net). Specifically, we develop original dynamic-snake convolution (DSConv) in CV and propose a novel DSConv module, which can extract the horizontal continuous feature in the shallow feature of the shot gather. Many experiments have shown that DSU-Net demonstrates higher accuracy and robustness than the other 2-D segmentation-based models, achieving state-of-the-art (SOTA) performance in 2-D seismic field surveys. Particularly, it can effectively detect FB jumps and better ensure the horizontal continuity of FB. In addition, the ablation experiment and the anti-noise experiment, respectively, verify the optimal structure of the DSConv module and the robustness of the picking.

Figures (11)

• Figure 1: A showcase of the 3$\times$3 convolution operation.
• Figure 2: Three types of convolution kernels. (a) Traditional convolution kernel. (b) Deformable convolution kernel. (c) X-direction dynamic snake convolution kernel. (d) Y-direction dynamic snake convolution kernel.
• Figure 3: The structure chart of DSU-Net. (a) The basic architecture of DSU-Net. In the first layer of the encoder, we utilize a stack of two DSU modules to deeply extract superficial features. (b) The specific structure of the DSU module. (c) The traditional convolution (TraConv) module. (d) Up-sampling module.
• Figure 4: Four classic shot gathers of Lalor, Sudbury, Brunswick, and Halfmile. The manual picking is marked on each sub-figure using the green curves.
• Figure 5: Comparison of quantitative results of deep learning-based picking methods on Fold #2.