Simultaneous reconstruction and denoising for DAS-VSP seismic data by RRU-net

Distributed acoustic sensing in vertical seismic profile (DAS-VSP) acquisition plays an important role reservoir monitoring. But the field data can be noisy and associated with missing traces which affects imaging geological interpretation. Therefore, DAS-VSP reconstruction a high signal-to-noise ratio (SNR) is worth studying. There are no exact relationships between signals noise t-x domain data, means that reconstructing suppressing simultaneously by deep neural network difficult. We develop novel algorithm based on U-net combination Hankel matrix as input/output, rather than data. The frequency of proposed to facilitate denoising rank reduction problem high-rank matrix. matrices incomplete ones while those complete without low-rank ones, beneficial learning. In our (RRU-net), two-channel input/output layers designed for real imaginary parts domain. Thus, reconstructed precision SNR could obtained using trained RRU-net. Meanwhile, we tested RRU-net two synthetic one results show effectiveness feasibility method. Our performs better both U-net-based method uses t?x approach.