Pub Date : 2025-03-21DOI: 10.1109/JOE.2025.3535563
Zhijie Tang;Yang Li;Chi Wang
In the realm of underwater detection technologies, reconstructing the three-dimensional structure of underwater objects is crucial for applications such as underwater target tracking, target locking, and navigational guidance. As a primary tool for underwater detection, acoustical imaging faces significant challenges in recovering the three-dimensional structure of objects from two-dimensional images. Current 3-D reconstruction methods mainly focus on reconstructing objects at the riverbed, overlooking the reconstruction of objects in the water in the absence of shadows. This study introduces a multiangle shape and height recovery method for such specific situations. By fixing the sonar detection angle and utilizing ViewPoint software to measure the contours of objects at different depths, a superimposition technique for two-dimensional sonar images was developed to achieve three-dimensional reconstruction of shadowless sonar image data. The proposed method is specifically designed for scenarios with diffuse echoes, where the sound waves scatter from rough surfaces rather than reflect specularly from smooth surfaces. This limitation ensures the method's applicability to objects lacking strong mirror-like reflections. This technique has been validated on three different categories of targets, with the reconstructed 3-D models accurately compared to the actual size and shape of the targets, demonstrating the method's effectiveness and providing a theoretical and methodological foundation for the 3-D reconstruction of underwater sonar targets.
{"title":"Multiangle Sonar Imaging for 3-D Reconstruction of Underwater Objects in Shadowless Environments","authors":"Zhijie Tang;Yang Li;Chi Wang","doi":"10.1109/JOE.2025.3535563","DOIUrl":"https://doi.org/10.1109/JOE.2025.3535563","url":null,"abstract":"In the realm of underwater detection technologies, reconstructing the three-dimensional structure of underwater objects is crucial for applications such as underwater target tracking, target locking, and navigational guidance. As a primary tool for underwater detection, acoustical imaging faces significant challenges in recovering the three-dimensional structure of objects from two-dimensional images. Current 3-D reconstruction methods mainly focus on reconstructing objects at the riverbed, overlooking the reconstruction of objects in the water in the absence of shadows. This study introduces a multiangle shape and height recovery method for such specific situations. By fixing the sonar detection angle and utilizing ViewPoint software to measure the contours of objects at different depths, a superimposition technique for two-dimensional sonar images was developed to achieve three-dimensional reconstruction of shadowless sonar image data. The proposed method is specifically designed for scenarios with diffuse echoes, where the sound waves scatter from rough surfaces rather than reflect specularly from smooth surfaces. This limitation ensures the method's applicability to objects lacking strong mirror-like reflections. This technique has been validated on three different categories of targets, with the reconstructed 3-D models accurately compared to the actual size and shape of the targets, demonstrating the method's effectiveness and providing a theoretical and methodological foundation for the 3-D reconstruction of underwater sonar targets.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"1344-1355"},"PeriodicalIF":3.8,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852452","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-21DOI: 10.1109/JOE.2025.3535591
Juan Li;Yajie Bai;Xuerong Cui;Lei Li;Bin Jiang;Shibao Li;Jungang Yang
Aiming at the problems of missing data and outliers in ocean observations and incomplete characterization of thermohaline related features, a 3-D thermohaline reconstruction model of the ocean based on multisource data are proposed. Multisource data from remote sensing and Current and Pressure recording Inverse Echo Sounders were used to analyze the projection relationship between 12-D features, such as sea surface temperature, bidirectional propagation time, and seafloor current velocity, and the distribution of ocean temperature and salinity at different depths (10–1000 m). A Bayesian optimization algorithmic framework is used to evaluate and gradually remove uncertainty from currently known data during the iterative process by extracting network parameters from the approximate probability distribution. More informed decision making improves the stability of the iterative process and reconstruction. In addition, a self-attention mechanism is introduced to dynamically focus on the dependencies between features of different dimensions by calculating the correlation matrix between features at arbitrary locations, enabling the model to more comprehensively characterize the thermohaline distribution and its changes. A Self-attentive Bayesian neural network (SABNN) model is established through empirical regression. The reconstructed model is validated using observational data from the Gulf of Mexico, and the experimental results show that the SABNN model has a significant improvement in temperature and salinity reconstruction accuracy compared with other network models or methods, with the RMSE and $R^{2}$ improved by more than 29.68%, 21.14% and 31.01%, 37.33%, respectively.
{"title":"Oceanic 3-D Thermohaline Field Reconstruction With Multidimensional Features Using SABNN","authors":"Juan Li;Yajie Bai;Xuerong Cui;Lei Li;Bin Jiang;Shibao Li;Jungang Yang","doi":"10.1109/JOE.2025.3535591","DOIUrl":"https://doi.org/10.1109/JOE.2025.3535591","url":null,"abstract":"Aiming at the problems of missing data and outliers in ocean observations and incomplete characterization of thermohaline related features, a 3-D thermohaline reconstruction model of the ocean based on multisource data are proposed. Multisource data from remote sensing and Current and Pressure recording Inverse Echo Sounders were used to analyze the projection relationship between 12-D features, such as sea surface temperature, bidirectional propagation time, and seafloor current velocity, and the distribution of ocean temperature and salinity at different depths (10–1000 m). A Bayesian optimization algorithmic framework is used to evaluate and gradually remove uncertainty from currently known data during the iterative process by extracting network parameters from the approximate probability distribution. More informed decision making improves the stability of the iterative process and reconstruction. In addition, a self-attention mechanism is introduced to dynamically focus on the dependencies between features of different dimensions by calculating the correlation matrix between features at arbitrary locations, enabling the model to more comprehensively characterize the thermohaline distribution and its changes. A Self-attentive Bayesian neural network (SABNN) model is established through empirical regression. The reconstructed model is validated using observational data from the Gulf of Mexico, and the experimental results show that the SABNN model has a significant improvement in temperature and salinity reconstruction accuracy compared with other network models or methods, with the RMSE and <inline-formula><tex-math>$R^{2}$</tex-math></inline-formula> improved by more than 29.68%, 21.14% and 31.01%, 37.33%, respectively.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"1273-1289"},"PeriodicalIF":3.8,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852391","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Autonomous underwater vehicles can carry multiple sensors, such as optical cameras and sonars, providing a common platform for underwater multimodal object detection. High-resolution optical images contain color information but are not applicable to turbid water environments. In contrast, acoustical waves are highly penetrating and travel long distances, making them suitable for low-light, turbid underwater environments, but sonar imaging has low resolution. The combination of the two can play to their respective advantages. This article presents a novel paradigm for underwater object detection using acousto–optic fusion (AO-UOD). Given that there is no publicly available data set, this article first constructs a paired data set for fusing optical and sonar images for underwater object detection. Paired sonar images and optical images were acquired by aligning the simulated plane of the ocean bottom. Based on this, a dual-stream interactive object detection network is designed. The network utilizes the structures of the fusion backbone, dual neck, and dual head to establish cross-modal information interaction between acoustical and optical. The attention interactive twin-branch fusion module is used to realize the fusion between features. Experimental results on the data collected show that AO-UOD can effectively fuse optical and sonar images to achieve robust detection performance. The multimodal method can utilize more information and possesses greater advantages over the unimodal method. This research not only provides a solid theoretical foundation for future multimodal object detection in marine environments but also points out the direction of technology development in practical applications.
{"title":"AO-UOD: A Novel Paradigm for Underwater Object Detection Using Acousto–Optic Fusion","authors":"Fengxue Yu;Fengqi Xiao;Congcong Li;En Cheng;Fei Yuan","doi":"10.1109/JOE.2025.3529121","DOIUrl":"https://doi.org/10.1109/JOE.2025.3529121","url":null,"abstract":"Autonomous underwater vehicles can carry multiple sensors, such as optical cameras and sonars, providing a common platform for underwater multimodal object detection. High-resolution optical images contain color information but are not applicable to turbid water environments. In contrast, acoustical waves are highly penetrating and travel long distances, making them suitable for low-light, turbid underwater environments, but sonar imaging has low resolution. The combination of the two can play to their respective advantages. This article presents a novel paradigm for underwater object detection using acousto–optic fusion (AO-UOD). Given that there is no publicly available data set, this article first constructs a paired data set for fusing optical and sonar images for underwater object detection. Paired sonar images and optical images were acquired by aligning the simulated plane of the ocean bottom. Based on this, a dual-stream interactive object detection network is designed. The network utilizes the structures of the fusion backbone, dual neck, and dual head to establish cross-modal information interaction between acoustical and optical. The attention interactive twin-branch fusion module is used to realize the fusion between features. Experimental results on the data collected show that AO-UOD can effectively fuse optical and sonar images to achieve robust detection performance. The multimodal method can utilize more information and possesses greater advantages over the unimodal method. This research not only provides a solid theoretical foundation for future multimodal object detection in marine environments but also points out the direction of technology development in practical applications.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"919-940"},"PeriodicalIF":3.8,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-21DOI: 10.1109/JOE.2025.3538954
Yan Liu;Yue Zhao;Bin Yu;Changsheng Zhu;Guanying Huo;Qingwu Li
With the development and utilization of marine resources, object detection in shallow sea environments becomes crucial. In real underwater environments, targets are often affected by motion blur or appear clustered, increasing detection difficulty. To address this problem, we propose an improved YOLOv8-based shallow sea creatures object detection method. We integrate receptive-field coordinate attention (RFCA) into the cross-stage partial bottleneck with the two convolutions (C2f) module of YOLOv8, creating the RFCA-enhanced C2f (C2f_RFCA). This enhancement improves feature extraction and fusion by leveraging multiscale receptive fields and refined feature fusion strategies, enabling better detection of blurred and occluded objects. The C2f_RFCA module captures both local and global features, enhancing detection accuracy in complex underwater scenarios. We additionally devised an improved dynamic head by substituting the deformable ConvNets version two (DCNv2) with DCNv3, forming dynamic head with DCNv3. This upgrade increases the flexibility of feature mapping and improves accuracy in detecting densely clustered objects by allowing adaptive receptive fields and enhancing boundary delineation. To evaluate the algorithm performance, we trained it on real-world underwater object detection data sets and conducted generalization experiments on detecting underwater objects, the underwater robot professional competition 2020 and underwater target detection and classification 2020 data sets. Experimental results show that, compared with YOLOv8n, our method increases mAP@0.5 by 1.9%, 1.7%, 4.3%, and 3.3%, and mAP@0.5:0.95 by 2.9%, 2.2%, 3.8%, and 5.0% in the four data sets. The proposed method significantly improves object detection accuracy for organisms in complex marine environments.
{"title":"An Improved YOLOv8-Based Shallow Sea Creatures Object Detection Method","authors":"Yan Liu;Yue Zhao;Bin Yu;Changsheng Zhu;Guanying Huo;Qingwu Li","doi":"10.1109/JOE.2025.3538954","DOIUrl":"https://doi.org/10.1109/JOE.2025.3538954","url":null,"abstract":"With the development and utilization of marine resources, object detection in shallow sea environments becomes crucial. In real underwater environments, targets are often affected by motion blur or appear clustered, increasing detection difficulty. To address this problem, we propose an improved YOLOv8-based shallow sea creatures object detection method. We integrate receptive-field coordinate attention (RFCA) into the cross-stage partial bottleneck with the two convolutions (C2f) module of YOLOv8, creating the RFCA-enhanced C2f (C2f_RFCA). This enhancement improves feature extraction and fusion by leveraging multiscale receptive fields and refined feature fusion strategies, enabling better detection of blurred and occluded objects. The C2f_RFCA module captures both local and global features, enhancing detection accuracy in complex underwater scenarios. We additionally devised an improved dynamic head by substituting the deformable ConvNets version two (DCNv2) with DCNv3, forming dynamic head with DCNv3. This upgrade increases the flexibility of feature mapping and improves accuracy in detecting densely clustered objects by allowing adaptive receptive fields and enhancing boundary delineation. To evaluate the algorithm performance, we trained it on real-world underwater object detection data sets and conducted generalization experiments on detecting underwater objects, the underwater robot professional competition 2020 and underwater target detection and classification 2020 data sets. Experimental results show that, compared with YOLOv8n, our method increases mAP@0.5 by 1.9%, 1.7%, 4.3%, and 3.3%, and mAP@0.5:0.95 by 2.9%, 2.2%, 3.8%, and 5.0% in the four data sets. The proposed method significantly improves object detection accuracy for organisms in complex marine environments.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"817-834"},"PeriodicalIF":3.8,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Light is absorbed, reflected, and refracted in an underwater environment due to the interaction between water and light. The red and blue channels in an image are attenuated due to these interactions. The red, green, and blue channels are typically employed as inputs for deep learning models, and the color casts, which result from different attenuation rates of the three channels, may affect the model's generalization performance. Besides, the color casts existing in the reference images will impact the deep-learning models. To address these challenges, a single channel network (SCN) model is introduced, which exclusively employs the green channel as its input, and is unaffected by the attenuations in the red and blue channels. An innovative feature processing module is presented, in which the characteristics of transformers and convolutional layers are fused to capture nonlinear relationships among the red, green, and blue channels. The public EUVP and LSUI data set experiments show that the proposed SCN model achieves competitive results with the existing best three channel models for the case of slight signal attenuation, and outperforms the existing state of arts three-channel models for the case of strong signal attenuation. Furthermore, the proposed model is trained on the self-built noncolor biased underwater image data set and is also tested on the public UCCS data set with three different types of color casts, whose experimental results exhibit balanced color distribution.
{"title":"SCN: A Novel Underwater Images Enhancement Method Based on Single Channel Network Model","authors":"Fuheng Zhou;Siqing Zhang;Yulong Huang;Pengsen Zhu;Yonggang Zhang","doi":"10.1109/JOE.2024.3474924","DOIUrl":"https://doi.org/10.1109/JOE.2024.3474924","url":null,"abstract":"Light is absorbed, reflected, and refracted in an underwater environment due to the interaction between water and light. The red and blue channels in an image are attenuated due to these interactions. The red, green, and blue channels are typically employed as inputs for deep learning models, and the color casts, which result from different attenuation rates of the three channels, may affect the model's generalization performance. Besides, the color casts existing in the reference images will impact the deep-learning models. To address these challenges, a single channel network (SCN) model is introduced, which exclusively employs the green channel as its input, and is unaffected by the attenuations in the red and blue channels. An innovative feature processing module is presented, in which the characteristics of transformers and convolutional layers are fused to capture nonlinear relationships among the red, green, and blue channels. The public EUVP and LSUI data set experiments show that the proposed SCN model achieves competitive results with the existing best three channel models for the case of slight signal attenuation, and outperforms the existing state of arts three-channel models for the case of strong signal attenuation. Furthermore, the proposed model is trained on the self-built noncolor biased underwater image data set and is also tested on the public UCCS data set with three different types of color casts, whose experimental results exhibit balanced color distribution.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"758-775"},"PeriodicalIF":3.8,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this article, we have experimentally demonstrated a laser Doppler velocimetry (3D-LDV) system capable of measuring 3-D flow velocities, employing a single emission wavelength and four photodetectors for capturing light scattered by particles in seawater. The optical measurement volume of the system is cylindrical and possesses dimensions that are significantly smaller than those of traditional acoustic Doppler systems—1.02 mm in diameter and 15.40 mm in length. This compact size renders the system particularly advantageous for applications demanding high spatial resolution, such as the observation of fine-scale turbulence. The performance of the 3D-LDV system was evaluated using a precision-controlled towing system in static seawater. It exhibited a measurement velocity range of 0.02–3.78 m/s, with a maximum relative error of 3.75%, a relative standard deviation of 1.49%, and an average directional angle deviation of 0.45° for angle changes within ±10°.
{"title":"Laser Doppler Velocimetry for 3-D Seawater Velocity Measurement Using a Single Wavelength","authors":"Lili Jiang;Xianglong Hao;Xinyu Zhang;Ran Song;Zhijun Zhang;Bingbing Li;Guangbing Yang;Xuejun Xiong;Juan Su;Chi Wu","doi":"10.1109/JOE.2025.3553941","DOIUrl":"https://doi.org/10.1109/JOE.2025.3553941","url":null,"abstract":"In this article, we have experimentally demonstrated a laser Doppler velocimetry (3D-LDV) system capable of measuring 3-D flow velocities, employing a single emission wavelength and four photodetectors for capturing light scattered by particles in seawater. The optical measurement volume of the system is cylindrical and possesses dimensions that are significantly smaller than those of traditional acoustic Doppler systems—1.02 mm in diameter and 15.40 mm in length. This compact size renders the system particularly advantageous for applications demanding high spatial resolution, such as the observation of fine-scale turbulence. The performance of the 3D-LDV system was evaluated using a precision-controlled towing system in static seawater. It exhibited a measurement velocity range of 0.02–3.78 m/s, with a maximum relative error of 3.75%, a relative standard deviation of 1.49%, and an average directional angle deviation of 0.45° for angle changes within ±10°.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 3","pages":"2200-2208"},"PeriodicalIF":3.8,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144646511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Deep learning (DL) approaches in underwater acoustic localization (UAL) have gained a great deal of popularity. While numerous works are devoted to improving the localization precision, they neglect another critical challenge inherent in the DL-based UAL problem, i.e., the model's practicality. Advanced DL models generally exhibit extremely high complexity, requiring a large amount of computational resources and resulting in slow inference time. Unfortunately, the limited processing power and real-time demands in oceanic applications make the deployment of complex DL models exceedingly challenging. To address this challenge, this article proposes a lightweight UAL framework based on knowledge distillation (KD) techniques, which effectively reduces the size of a deep UAL model while maintaining competitive performance. Specifically, a dedicated teacher network is designed using attention mechanisms and convolutional neural networks (CNNs). Then, the KD is performed to distill the knowledge from the teacher network into a lightweight student model, such as a three-layer CNN. In practical deployment, only the lightweight student model will be utilized. With the proposed lightweight framework, the student model has 98.68% fewer model parameters and is 87.4% faster in inference time compared to the teacher network, while the prediction accuracy drops to only 1.07% (97.55% $rightarrow$ 96.48%). In addition, the generalization ability of the student model is examined through transfer learning, where the model is transferred between two different ocean environments. The student model demonstrates a stronger generalization ability compared to the model without the KD process, as it can quickly adapt itself to a new application environment using just 10% of the data.
{"title":"Toward Real-World Applicability: Lightweight Underwater Acoustic Localization Model Through Knowledge Distillation","authors":"Runze Hu;Xiaohui Chu;Daowei Dou;Xiaogang Liu;Yining Liu;Bingbing Qi","doi":"10.1109/JOE.2025.3538928","DOIUrl":"https://doi.org/10.1109/JOE.2025.3538928","url":null,"abstract":"Deep learning (DL) approaches in underwater acoustic localization (UAL) have gained a great deal of popularity. While numerous works are devoted to improving the localization precision, they neglect another critical challenge inherent in the DL-based UAL problem, i.e., the model's practicality. Advanced DL models generally exhibit extremely high complexity, requiring a large amount of computational resources and resulting in slow inference time. Unfortunately, the limited processing power and real-time demands in oceanic applications make the deployment of complex DL models exceedingly challenging. To address this challenge, this article proposes a lightweight UAL framework based on knowledge distillation (KD) techniques, which effectively reduces the size of a deep UAL model while maintaining competitive performance. Specifically, a dedicated teacher network is designed using attention mechanisms and convolutional neural networks (CNNs). Then, the KD is performed to distill the knowledge from the teacher network into a lightweight student model, such as a three-layer CNN. In practical deployment, only the lightweight student model will be utilized. With the proposed lightweight framework, the student model has 98.68% fewer model parameters and is 87.4% faster in inference time compared to the teacher network, while the prediction accuracy drops to only 1.07% (97.55% <inline-formula><tex-math>$rightarrow$</tex-math></inline-formula> 96.48%). In addition, the generalization ability of the student model is examined through transfer learning, where the model is transferred between two different ocean environments. The student model demonstrates a stronger generalization ability compared to the model without the KD process, as it can quickly adapt itself to a new application environment using just 10% of the data.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"1429-1442"},"PeriodicalIF":3.8,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-18DOI: 10.1109/JOE.2024.3521256
David W. Caress;Eric J. Martin;Michael Risi;Giancarlo Troni;Andrew Hamilton;Chad Kecy;Jennifer B. Paduan;Hans J. Thomas;Stephen M. Rock;Monica Wolfson-Schwehr;Richard Henthorn;Brett Hobson;Larry E. Bird
The Monterey Bay Aquarium Research Institute has developed a low-altitude survey system (LASS) to conduct cm-scale seafloor surveys of complex terrain in the deep ocean. The LASS is integrated with a remotely operated vehicle (ROV), which is operated at a 3-m standoff to obtain 5-cm-lateral-resolution bathymetry using a multibeam sonar, 1-cm-resolution bathymetry using a wide-swath lidar laser scanner, and 2-mm/pixel resolution color photography using stereo still cameras illuminated by strobes. Surveys are typically conducted with 3-m line spacing and 0.2-m/s speed and executed autonomously by the ROV. The instrument frame actively rotates to keep the sensors oriented normal to the seafloor. The strobe lights, mounted on swing arms on either side of the ROV, similarly rotate to face the seafloor. Areas of 120 m × 120 m can be covered in about 8 h. Example surveys include 1) deep-sea soft coral and sponge communities from Sur Ridge, offshore Central California; 2) a warm venting site hosting thousands of brooding octopus near Davidson Seamount, also offshore Central California; and 3) a high-temperature hydrothermal vent field on Axial Seamount, on the Juan de Fuca Ridge. An advantage of combining optical and acoustic remote sensing is that the lidar and cameras map soft animals, while the multibeam sonar maps the solid seafloor. The long-term goal is to field these sensors from a hover-capable autonomous platform rather than ROVs, enabling efficient 1-cm-scale seafloor surveys in the deep ocean.
{"title":"The MBARI Low-Altitude Survey System for 1-cm-Scale Seafloor Surveys in the Deep Ocean","authors":"David W. Caress;Eric J. Martin;Michael Risi;Giancarlo Troni;Andrew Hamilton;Chad Kecy;Jennifer B. Paduan;Hans J. Thomas;Stephen M. Rock;Monica Wolfson-Schwehr;Richard Henthorn;Brett Hobson;Larry E. Bird","doi":"10.1109/JOE.2024.3521256","DOIUrl":"https://doi.org/10.1109/JOE.2024.3521256","url":null,"abstract":"The Monterey Bay Aquarium Research Institute has developed a low-altitude survey system (LASS) to conduct cm-scale seafloor surveys of complex terrain in the deep ocean. The LASS is integrated with a remotely operated vehicle (ROV), which is operated at a 3-m standoff to obtain 5-cm-lateral-resolution bathymetry using a multibeam sonar, 1-cm-resolution bathymetry using a wide-swath lidar laser scanner, and 2-mm/pixel resolution color photography using stereo still cameras illuminated by strobes. Surveys are typically conducted with 3-m line spacing and 0.2-m/s speed and executed autonomously by the ROV. The instrument frame actively rotates to keep the sensors oriented normal to the seafloor. The strobe lights, mounted on swing arms on either side of the ROV, similarly rotate to face the seafloor. Areas of 120 m × 120 m can be covered in about 8 h. Example surveys include 1) deep-sea soft coral and sponge communities from Sur Ridge, offshore Central California; 2) a warm venting site hosting thousands of brooding octopus near Davidson Seamount, also offshore Central California; and 3) a high-temperature hydrothermal vent field on Axial Seamount, on the Juan de Fuca Ridge. An advantage of combining optical and acoustic remote sensing is that the lidar and cameras map soft animals, while the multibeam sonar maps the solid seafloor. The long-term goal is to field these sensors from a hover-capable autonomous platform rather than ROVs, enabling efficient 1-cm-scale seafloor surveys in the deep ocean.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 3","pages":"1573-1584"},"PeriodicalIF":3.8,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10931848","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144646563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-18DOI: 10.1109/JOE.2025.3535597
Cong Peng;Lei Wang;Juncheng Gao;Shuhao Zhang;Haoran Ji
This article proposes a new active sonar detector based on a beamformed deep neural network (BDNN) in three steps. The process involves a preprocessing step, a deep neural network (DNN) application step, and a subsequent postprocessing step. In the preprocessing step, partial spectra are extracted from multiple directions through frequency-domain beamforming. These partial spectra from different directions serve as DNN input, yielding estimated target probabilities as output in the DNN application step. In the postprocessing step, a multiframe probability multiplication technique is proposed, and the number of frames is determined adaptively. The proposed BDNN generates a gridded azimuth-distance graph, where each grid cell represents the probability of a target's presence at a specific azimuth and distance. To guarantee real-time application, we also propose a graphics processing unit based parallel acceleration method, which increases the computation speed of the beamforming process by nearly two orders of magnitude compared to the CPU. The proposed BDNN is verified through sea and lake trials. The results demonstrate that the proposed BDNN achieves better detection performance compared to the conventional matched filter method and exhibits remarkable generalization capabilities.
{"title":"A New Active Sonar Detector Based on Beamformed Deep Neural Network","authors":"Cong Peng;Lei Wang;Juncheng Gao;Shuhao Zhang;Haoran Ji","doi":"10.1109/JOE.2025.3535597","DOIUrl":"https://doi.org/10.1109/JOE.2025.3535597","url":null,"abstract":"This article proposes a new active sonar detector based on a beamformed deep neural network (BDNN) in three steps. The process involves a preprocessing step, a deep neural network (DNN) application step, and a subsequent postprocessing step. In the preprocessing step, partial spectra are extracted from multiple directions through frequency-domain beamforming. These partial spectra from different directions serve as DNN input, yielding estimated target probabilities as output in the DNN application step. In the postprocessing step, a multiframe probability multiplication technique is proposed, and the number of frames is determined adaptively. The proposed BDNN generates a gridded azimuth-distance graph, where each grid cell represents the probability of a target's presence at a specific azimuth and distance. To guarantee real-time application, we also propose a graphics processing unit based parallel acceleration method, which increases the computation speed of the beamforming process by nearly two orders of magnitude compared to the CPU. The proposed BDNN is verified through sea and lake trials. The results demonstrate that the proposed BDNN achieves better detection performance compared to the conventional matched filter method and exhibits remarkable generalization capabilities.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 2","pages":"1370-1386"},"PeriodicalIF":3.8,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143852432","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-16DOI: 10.1109/JOE.2025.3550985
Charles-Antoine Guérin
We describe and exploit a reformulation, based on a recently introduced change of variables, of the double integral that describes the second-order ocean Doppler spectrum measured by high-frequency radars. We show that this alternative expression, which was primarily designed for improving the numerical inversion of the ocean wave spectrum, is also advantageous for the analytical inversion of the main sea state parameters. To this end, we revisit Barrick's Method for the estimation of the significant wave height and the mean period from the ocean Doppler spectrum. On the basis of numerical simulations we show that a better estimation of these parameters can be achieved, which necessitates a preliminary bias correction that depends only on the radar frequency. A second consequence of this improved formulation is the derivation of a simple yet analytical nonlinear approximation of the second-order ocean Doppler spectrum when the Doppler frequency is larger than the Bragg frequency. This opens up new perspectives for the inversion of directional wave spectra from high-frequency radar measurements.
{"title":"Improved Calculation of the Second-Order Ocean Doppler Spectrum for Sea State Inversion","authors":"Charles-Antoine Guérin","doi":"10.1109/JOE.2025.3550985","DOIUrl":"https://doi.org/10.1109/JOE.2025.3550985","url":null,"abstract":"We describe and exploit a reformulation, based on a recently introduced change of variables, of the double integral that describes the second-order ocean Doppler spectrum measured by high-frequency radars. We show that this alternative expression, which was primarily designed for improving the numerical inversion of the ocean wave spectrum, is also advantageous for the analytical inversion of the main sea state parameters. To this end, we revisit Barrick's Method for the estimation of the significant wave height and the mean period from the ocean Doppler spectrum. On the basis of numerical simulations we show that a better estimation of these parameters can be achieved, which necessitates a preliminary bias correction that depends only on the radar frequency. A second consequence of this improved formulation is the derivation of a simple yet analytical nonlinear approximation of the second-order ocean Doppler spectrum when the Doppler frequency is larger than the Bragg frequency. This opens up new perspectives for the inversion of directional wave spectra from high-frequency radar measurements.","PeriodicalId":13191,"journal":{"name":"IEEE Journal of Oceanic Engineering","volume":"50 3","pages":"1895-1905"},"PeriodicalIF":3.8,"publicationDate":"2025-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144646460","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：