Horizon Picking from SBP Images Using Physicals-Combined Deep Learning

被引:2
作者
Feng, Jie [1 ,2 ]
Zhao, Jianhu [1 ,2 ]
Zheng, Gen [1 ,2 ]
Li, Shaobo [1 ,2 ]
机构
[1] Wuhan Univ, Sch Geodesy & Geomat, Wuhan 430079, Peoples R China
[2] Wuhan Univ, Inst Marine Sci & Technol, Wuhan 430079, Peoples R China
基金
中国国家自然科学基金; 国家重点研发计划;
关键词
sub-bottom profiler; horizon picking; deep learning; multiple suppression; CHIRP SUBBOTTOM PROFILER; SITE;
D O I
10.3390/rs13183565
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Horizon picking from sub-bottom profiler (SBP) images has great significance in marine shallow strata studies. However, the mainstream automatic picking methods cannot handle multiples well, and there is a need to set a group of parameters manually. Considering the constant increase in the amount of SBP data and the high efficiency of deep learning (DL), we proposed a physicals-combined DL method to pick the horizons from SBP images. We adopted the DeeplabV3+ net to extract the horizons and multiples from SBP images. We generated a training dataset from the Jiaozhou Bay survey (Shandong, China) and the Zhujiang estuary survey (Guangzhou, China) to increase the applicability of the trained model. After the DL processing, we proposed a simulated Radon transform method to eliminate the surface-related multiples from the prediction by combining the designed pseudo-Radon transform and correlation analysis. We verified the proposed method using actual data (not involved in the training dataset) from Jiaozhou Bay and Zhujiang estuary. The positions of picked horizons are accurate, and multiples are suppressed.
引用
收藏
页数:22
相关论文
共 34 条
[1]   Imaging of multiple reflections [J].
Berkhout, A. J. ;
Verschuur, D. J. .
GEOPHYSICS, 2006, 71 (04) :SI209-SI220
[2]   SEISMIC HORIZON DETECTION USING IMAGE-PROCESSING ALGORITHMS [J].
BONDAR, I .
GEOPHYSICAL PROSPECTING, 1992, 40 (07) :785-800
[3]   Illuminating centimeter-level resolution stratum via developed high-frequency sub-bottom profiler mounted on Deep-Sea Warrior deep-submergence vehicle [J].
Cao, Xinghui ;
Qu, Zhiguo ;
Shen, Binjian ;
Zhang, Hanyu .
MARINE GEORESOURCES & GEOTECHNOLOGY, 2021, 39 (11) :1296-1306
[4]   Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation [J].
Chen, Liang-Chieh ;
Zhu, Yukun ;
Papandreou, George ;
Schroff, Florian ;
Adam, Hartwig .
COMPUTER VISION - ECCV 2018, PT VII, 2018, 11211 :833-851
[5]   Structured Object-Level Relational Reasoning CNN-Based Target Detection Algorithm in a Remote Sensing Image [J].
Cheng, Bei ;
Li, Zhengzhou ;
Xu, Bitong ;
Yao, Xu ;
Ding, Zhiquan ;
Qin, Tianqi .
REMOTE SENSING, 2021, 13 (02) :1-27
[6]   Evidence of extensive carbonate mounds and sublacustrine channels in shallow waters of Lake Van, eastern Turkey, based on high-resolution chirp subbottom profiler and multibeam echosounder data [J].
Cukur, Deniz ;
Krastel, Sebastian ;
Cagatay, M. Namik ;
Damci, Emre ;
Meydan, Aysegul Feray ;
Kim, Seong-Pil .
GEO-MARINE LETTERS, 2015, 35 (05) :329-340
[7]  
[丁维凤 DING Weifeng], 2008, [海洋科学, Marine Sciences], V32, P1
[8]   Automated reflection picking and polarity assessment through attribute analysis: Theory and application to synthetic and real ground-penetrating radar data [J].
Dossi, Matteo ;
Forte, Emanuele ;
Pipan, Michele .
GEOPHYSICS, 2015, 80 (05) :H23-H35
[9]   Multi-Frequency, Multi-Sonar Mapping of Shallow HabitatsEfficacy and Management Implications in the National Marine Park of Zakynthos, Greece [J].
Fakiris, Elias ;
Blondel, Philippe ;
Papatheodorou, George ;
Christodoulou, Dimitris ;
Dimas, Xenophon ;
Georgiou, Nikos ;
Kordella, Stavroula ;
Dimitriadis, Charalampos ;
Rzhanov, Yuri ;
Geraga, Maria ;
Ferentinos, George .
REMOTE SENSING, 2019, 11 (04)
[10]   Detection of Boulders in Side Scan Sonar Mosaics by a Neural Network [J].
Feldens, Peter ;
Darr, Alexander ;
Feldens, Agata ;
Tauber, Franz .
GEOSCIENCES, 2019, 9 (04)