Deep Prior-Based Unsupervised Reconstruction of Irregularly Sampled Seismic Data

Release time:2026-08-11  Hits:

• DOI number:10.1109/LGRS.2020.3044455
• Affiliation of Author(s):中国石油大学（华东）;米兰理工大学
• Teaching and Research Group:地球科学与技术学院;电子信息系
• Journal:IEEE Geoscience and Remote Sensing Letters
• Place of Publication:美国
• Funded by:国家留学基金委;中央高校基本科研业务费
• Key Words:seismic data interpolation; deep prior; unsupervised reconstruction; 3D U-Net
• Abstract:提出一种基于深度先验的无监督地震数据插值方法，以多分辨率三维U-Net作为隐式先验，不依赖预训练样本即可求解不规则采样数据的重建问题。合成和现场数据实验验证了该方法对缺失地震道的有效恢复能力。
• Note:来源：https://doi.org/10.1109/LGRS.2020.3044455
• Co-author:Francesco Picetti, Vincenzo Lipari, Paolo Bestagini, Xiaoming Tang, Stefano Tubaro
• First Author:Fantong Kong
• Indexed by:科研论文
• Discipline:Natural Science
• First-Level Discipline:Geophysics
• Document Type:期刊论文
• Volume:19
• Page Number:1-5
• ISSN No.:1545-598X;1558-0571
• Date of Publication:2022-01-01