Reverse time migration (RTM) is a technique used to obtain high-resolution images of underground reflectors; however, this method is computationally intensive when dealing with large amounts of seismic data. Multisource RTM can significantly reduce the computational cost by processing multiple shots simultaneously. However, multisource-based methods frequently result in crosstalk artifacts in the migrated images, causing serious interference in the imaging signals. Plane-wave migration, as a mainstream multisource method, can yield migrated images with plane waves in different angles by implementing phase encoding of the source and receiver wavefields; however, this method frequently requires a trade-off between computational efficiency and imaging quality. We propose a method based on deep learning for removing crosstalk artifacts and enhancing the image quality of plane-wave migration images. We designed a convolutional neural network that accepts an input of seven plane-wave images at different angles and outputs a clear and enhanced image. We built over 500 1024×2561024\,\times \, 256 velocity models, and employed each of them using plane-wave migration to produce raw images at 0°, ±10°, ±20°, and ±30° as input of the network. Labels are high-resolution images computed from the corresponding reflectivity models by convolving with a Ricker wavelet. Random sub-images with a size of 512×128512\,\times \, 128 were used for training the network. Numerical examples demonstrated the effectiveness of the trained network in crosstalk removal and imaging enhancement. The proposed method is superior to both the conventional RTM and plane-wave RTM (PWRTM) in imaging resolution. Moreover, the proposed method requires only seven migrations, significantly improving the computational efficiency. In the numerical examples, the processing time required by our method was approximately 1.6% and 10% of that required by RTM and PWRTM, respectively.

中文翻译：

---

深度学习增强多源逆时偏移


逆时偏移（RTM）是一种用于获取地下反射体高分辨率图像的技术；然而，在处理大量地震数据时，该方法计算量较大。多源 RTM 可以通过同时处理多个镜头来显着降低计算成本。然而，基于多源的方法经常会导致迁移图像中出现串扰伪影，从而对成像信号造成严重干扰。平面波偏移作为主流的多源方法，通过对源波场和接收波场进行相位编码，可以得到不同角度平面波的偏移图像；然而，这种方法经常需要在计算效率和成像质量之间进行权衡。我们提出了一种基于深度学习的方法，用于消除串扰伪影并提高平面波偏移图像的图像质量。我们设计了一个卷积神经网络，它接受七个不同角度的平面波图像的输入，并输出清晰且增强的图像。我们建立了超过 500 个 1024×2561024\,\times \, 256 个速度模型，并使用每个模型使用平面波偏移来生成 0°、±10°、±20° 和 ±30° 的原始图像，作为网络。标签是通过与 Ricker 小波卷积从相应的反射率模型计算出的高分辨率图像。使用大小为 512×128512\,\times \, 128 的随机子图像来训练网络。数值例子证明了经过训练的网络在串扰消除和成像增强方面的有效性。该方法在成像分辨率方面优于传统RTM和平面波RTM（PWRTM）。 此外，该方法仅需要七次迁移，显着提高了计算效率。在数值示例中，我们的方法所需的处理时间分别约为 RTM 和 PWRTM 所需处理时间的 1.6% 和 10%。