TY - JOUR
T1 - Joint Physics and Data Driven Full-Waveform Inversion for Underground Dielectric Targets Imaging
AU - Sun, Haoran
AU - Yang, Xiaopeng
AU - Gong, Junbo
AU - Qu, Xiaodong
AU - Lan, Tian
N1 - Publisher Copyright:
© 1980-2012 IEEE.
PY - 2022
Y1 - 2022
N2 - To better reconstruct underground targets based on ground-penetrating radar (GPR) data, this article proposes a joint physics and data driven full-waveform inversion (PDD-FWI) scheme. This scheme combines a physics-based noniterative approach and a data driven deep neural network (DNN) to reconstruct target location, shape, and permittivity accurately. First, the normalized range migration algorithm (RMA) is introduced to extract the target contour and location information, which not only improves the robustness of the proposed scheme but also ensures adaptability to different GPR equipment. Then, the GPR dielectric target reconstruction network (GPRDtrNet) is developed based on the improved U-net structure, including reducing network layers and adding multiscale additive spatial attention gates and skip-connection structures. Compared with previous DNN-based reconstruction methods, GPRDtrNet has the advantages of small data requirement, high accuracy, strong generalization, and noise tolerance. Finally, the simulated and real dataset containing kinds of targets is constructed to train and test GPRDtrNet. The results show that the proposed method can reconstruct underground dielectric targets accurately with high robustness and noise tolerance.
AB - To better reconstruct underground targets based on ground-penetrating radar (GPR) data, this article proposes a joint physics and data driven full-waveform inversion (PDD-FWI) scheme. This scheme combines a physics-based noniterative approach and a data driven deep neural network (DNN) to reconstruct target location, shape, and permittivity accurately. First, the normalized range migration algorithm (RMA) is introduced to extract the target contour and location information, which not only improves the robustness of the proposed scheme but also ensures adaptability to different GPR equipment. Then, the GPR dielectric target reconstruction network (GPRDtrNet) is developed based on the improved U-net structure, including reducing network layers and adding multiscale additive spatial attention gates and skip-connection structures. Compared with previous DNN-based reconstruction methods, GPRDtrNet has the advantages of small data requirement, high accuracy, strong generalization, and noise tolerance. Finally, the simulated and real dataset containing kinds of targets is constructed to train and test GPRDtrNet. The results show that the proposed method can reconstruct underground dielectric targets accurately with high robustness and noise tolerance.
KW - Dielectric targets- reconstruction
KW - ground-penetrating radar (GPR)
KW - multiscale attention gates
KW - neural network
UR - http://www.scopus.com/inward/record.url?scp=85141633493&partnerID=8YFLogxK
U2 - 10.1109/TGRS.2022.3219138
DO - 10.1109/TGRS.2022.3219138
M3 - Article
AN - SCOPUS:85141633493
SN - 0196-2892
VL - 60
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
M1 - 4513311
ER -