MSSPN: Automatic First Arrival Picking using Multi-Stage Segmentation-Picking Network

Abstract

Picking the first arrival of prestack gathers is an indispensable step in seismic data processing. To enhance the efficiency of seismic data processing, some deep-learning-based methods for first arrival picking have been proposed. However, when applying currently trained models to data that significantly differs from the training set, the results are often suboptimal. We refer to this predictive scenario as cross-survey picking. Therefore, further improving model generalization for accurate cross-survey picking has become an urgent problem. To overcome the problem, we propose a multi-stage picking method named Multi-Stage Segmentation-Picking Network (MSSPN), which breaks down the complex picking task into four stages. In the first stage, we propose a Coarse Segmentation Network (CSN) to recognize a rough trend of first arrivals. Second, a robust trend estimation method is proposed in the second stage to further obtain a tighter range of first arrivals. Third, a Refined Segmentation Network (RSN) is conducted in the third stage to pick high-precision first arrivals. Finally, we propose a velocity constraint-based post-processing strategy to remove the outliers of network pickings. Extensive experiments show that MSSPN outperforms current state-of-the-art methods under the cross-survey test situation in terms of the metrics of accuracy and stability. Particularly, MSSPN achieves 94.64% and 89.74% accuracy under the cross-survey field cases of the median and low signal-noise ratio (SNR) data, respectively.