Related references
Note: Only part of the references are listed.
Article
Geochemistry & Geophysics
Liang Zhang et al.
Summary: This letter presents an unsupervised ship detection method in SAR images using superpixel segmentation and cross stage partial network (CSPNet). The method achieves pixel-level detection map instead of bounding box result.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Geochemistry & Geophysics
Yin Zhuang et al.
Summary: Under the multiscale distribution, a concise and effective one-stage anchor-free contour modeling detector called CMDet is proposed for accurate arbitrary-oriented ship detection. Different from currently existed methods, CMDet resolves the oriented bounding box (OBB) modeling by jointly regressing the contour information. The proposed CMDet achieves competitive results compared to state-of-the-art detectors in extensive experiments on public OBB ship detection datasets.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Engineering, Electrical & Electronic
Bo Guo et al.
Summary: Fine-grained ship detection is a challenging task due to large aspect ratios and severe category imbalance. To address this, we propose a shape-aware feature learning method to mitigate misalignments and a shape-aware instance switching method to balance category distribution. Experimental results on our multicategory ship detection dataset demonstrate the superiority of our proposed method over state-of-the-art methods.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Engineering, Electrical & Electronic
Lin Bai et al.
Summary: This article proposes a novel SAR ship detection network called FEPS-Net, which aims to solve the challenges in ship detection in optical remote sensing images. The network utilizes a feature enhancement pyramid and a shallow feature reconstruction module to enhance weak signals and detect small ships. Experimental results demonstrate the advantages of FEPS-Net in SAR ship detection.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Engineering, Aerospace
Zhaocheng Wang et al.
Summary: Due to the large sizes of SAR images, traditional sliding window-based ship detection methods suffer from high computation redundancy and numerous false alarms. To address this issue, a novel ship detection method called global and local context-aware ship detector is proposed. This method utilizes global and local context information to select necessary subimages and suppress false alarms. Experimental results demonstrate that the proposed method outperforms traditional deep learning methods in terms of precision and efficiency.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2023)
Article
Geochemistry & Geophysics
Yuang Du et al.
Summary: Target detection methods based on deep learning have made significant progress in SAR ship detection. However, obtaining target-level annotations for SAR images is challenging. To address this, we propose a semi-supervised SAR ship detection network that leverages scene-level annotations to improve detection performance. The network consists of a scene characteristic learning branch and a detection branch, with designed losses to utilize scene-level annotations. We also introduce a hierarchical test process that reduces false alarms and improves detection performance by considering different scene types. Experimental results demonstrate the effectiveness of our approach using two measured SAR ship detection datasets.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Tao Zhang et al.
Summary: In recent decades, the detection of ships using polarimetric synthetic aperture radar (PolSAR) has been a hot topic. However, most existing methods fail to detect small ships with weak backscattering. To address this issue, a ship detection matrix called complete polarimetric covariance matrix (CP) was proposed, but its calculation strategy still needs improvement. To overcome these limitations, an information reconstruction-based polarimetric covariance matrix (IC) is developed. Experimental results on GF-3 PolSAR datasets showed that the proposed IC matrix has better performance in ship detection, especially for small ships.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Computer Science, Information Systems
Wenxiao Zhao et al.
Summary: In this study, a new Artificial-SAR-Vessel dataset was generated by combining the FUSAR-Ship dataset and the SimpleCopyPaste method. A novel multi-category vessel detection method called CRAS-YOLO was proposed, which integrated a convolutional block attention module (CBAM), receptive fields block (RFB), and adaptively spatial feature fusion (ASFF) based on YOLOv5s. The experiments demonstrated that the proposed CRAS-YOLO model achieved high precision, recall rate, and mean average precision (mAP) (0.5) of up to 90.4%, 88.6%, and 92.1% respectively.
Article
Engineering, Electrical & Electronic
Hicham Madjidi et al.
Summary: In this study, an automatic bilateral censoring and detection method is proposed and analyzed for log-normal sea clutter using the AML-CFAR detector. By resorting to linear biparametric adaptive thresholds, a logarithmic amplifier is introduced to transform the distribution to Gaussian. The AML estimates of the unknown mean and standard deviation parameters are used to compute the censoring thresholds and estimate the detection threshold, resulting in better performance compared to state-of-the-art detectors in simulations on both simulated and real SAR images.
DIGITAL SIGNAL PROCESSING
(2023)
Article
Geography, Physical
Siyuan Zhao et al.
Summary: In this article, a domain adaptation (DA) Transformer object detection method is proposed to solve the unlabeled multisource satellite-borne SAR image object detection problem. The proposed method utilizes Vision Transformer (ViT) Faster Region CNN (FRCNN) as the baseline network to extract global features of SAR images and achieves the highest object detection accuracy compared to other state of the art (SOTA) methods. Furthermore, the method improves the accuracy by refining and reconstructing the pseudo-label of the target domain through feature clustering, and reduces the training time by more than 16% when compared to recently proposed Transformer-based methods.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2023)
Article
Energy & Fuels
Quan Jiang et al.
Summary: This paper proposes a multi-ship detection and tracking method based on YOLOv5x combined with DeepSort algorithm to solve the problem of low detection rate of silicon energy bulk material cargo ship targets in river monitoring videos. The experimental results show that this method can effectively supervise the illegal mining and transportation of silicon energy bulk material.
Article
Engineering, Electrical & Electronic
Jianwei Li et al.
Summary: Deep learning has played a crucial role in the development of synthetic aperture radar (SAR) ship detection. However, the heavy and computation intensive nature of the detectors hinders their usage on the edge. To overcome this issue, lightweight networks and acceleration ideas have been proposed. This survey reviews real-time SAR ship detection papers, covering model compression, acceleration methods, and various object detection techniques. It provides an overview of 70 papers, including years, datasets, journals, deep-learning frameworks, and hardware, as well as experimental results showing significant improvements in speed and accuracy of the algorithms.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Yunpeng Zhang et al.
Summary: Most existing multiobject tracking (MOT) algorithms focus on optical image datasets, but the synthetic aperture radar (SAR) image dataset presents challenges such as limited prior samples, high false alarm rate, and defocusing interference. We propose a robust MOT algorithm for multi-ship tracking in complex imaging conditions. The algorithm modifies the kernelized correlation filters (KCFs) algorithm to address false alarms, uses adaptive matching strategies based on intersection patterns to handle deviated detections, and introduces a tracker's time limit with Gaussian distribution to improve reassociation ability after defocusing interruptions. Experimental results demonstrate the algorithm's robust tracking performance.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Geochemistry & Geophysics
Weichang Zhang et al.
Summary: Automatic detection and localization of objects in remote sensing images are important for remote sensing systems. However, existing frameworks usually suffer from poor performance due to a lack of large-scale training datasets. To address this, a novel sensor-related image synthesis framework called RS-ISP is developed. It introduces designs to ensure distribution consistency between generated and real images, resulting in improved ship detection performance.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Linping Zhang et al.
Summary: This article introduces a novel deep learning network called YOLO-FA for SAR ship detection, which suppresses sea clutter by processing frequency-domain information. The proposed method achieves state-of-the-art detection performance on both high-resolution SAR image dataset and SAR ship detection dataset.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Review
Engineering, Civil
Sarang Thombre et al.
Summary: Autonomous ships, using perception systems and AI techniques, are expected to enhance safety and efficiency in maritime navigation. This article introduces the operational requirements for autonomous vessels and discusses suitable sensors and AI techniques for their perception systems. The integration of four sensor families and sources of auxiliary data are explained. The perception tasks involve problems that can be solved using AI techniques such as deep learning and Gaussian processes.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
Jiaying Lin et al.
Summary: Environment perception is crucial for automated maritime vehicles, but deep-learning-based object detection using LiDAR has not seen the same level of development in the maritime field as in the automotive sector. To address this, we propose a novel concept that uses LiDAR as the primary sensor and assisted by the automatic identification system (AIS) for maritime environment perception. Our approach includes object detection, multi-object tracking, and static environment mapping, demonstrating promising results in simulations and real-world evaluations.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Xueqian Wang et al.
Summary: This paper proposes a new method for fusion of spaceborne and airborne SAR images based on the target proposal and the copula theory (TPCT), which improves the target-to-clutter ratios of composite images and enhances ship detection performance. Experiment results using measured SAR data show that the TPCT fusion method outperforms other commonly used image fusion methods in ship detection tasks.
INFORMATION FUSION
(2022)
Article
Computer Science, Hardware & Architecture
Jose Escorcia-Gutierrez et al.
Summary: This paper presents an efficient optimal mask regional convolutional neural network technique for small ship detection. It resolves issues of limited real-world samples and achieves high accuracy through data augmentation and hyperparameter tuning.
COMPUTERS & ELECTRICAL ENGINEERING
(2022)
Article
Instruments & Instrumentation
Wenying Mo et al.
Summary: This paper proposes a TCS-SMoLGDR algorithm for nighttime infrared ship target detection. The algorithm takes advantage of the largest grayscale dynamic range of infrared ship targets at night and generates a saliency map through SMoLGDR. The real target area is determined using the connected domain mean strategy. To better separate targets with uneven grayscale distribution from the background, this paper proposes the TCS method. Experimental results demonstrate that the proposed algorithm can effectively detect small dim targets in nighttime infrared maritime images with higher accuracy compared to other algorithms.
INFRARED PHYSICS & TECHNOLOGY
(2022)
Article
Computer Science, Hardware & Architecture
Xin Lou et al.
Summary: This paper proposes a generative transfer learning framework for ship detection in Synthetic Aperture Radar (SAR) images by addressing the data acquisition and labeling problem. Experimental results show that the generated pseudo-SAR images can improve the generalization performance of detection models and reduce missed detections and false positives in complex backgrounds.
COMPUTERS & ELECTRICAL ENGINEERING
(2022)
Proceedings Paper
Computer Science, Theory & Methods
Parneet Kaur et al.
Summary: Vessels play a crucial role in international cargo transportation and the marine economy. However, the marine industry has not fully embraced modern technology to protect the blue economy. Intelligent vessel systems can enhance safety and decision-making for mariners, but advanced perception technology requires real-world data. The SeaSAw dataset is the largest maritime dataset, consisting of millions of images and objects, providing valuable resources for object detection, classification, and tracking in the marine domain.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2022
(2022)
Article
Environmental Sciences
Hichem Mahgoun et al.
Summary: Ships remote sensing is a crucial tool for marine traffic management. This research aims to identify the best ship detection procedure by combining Singular Values Decomposition (SVD) and Constant False Alarm Rate (CFAR) algorithms. The results show that the combination of SVD and CFAR algorithm based on the exponential distribution outperforms other methods, achieving a best probability of detection (PD) of 93% with a probability of false alarm (PFA) of 0.1%.
REMOTE SENSING APPLICATIONS-SOCIETY AND ENVIRONMENT
(2022)
Review
Engineering, Aerospace
Bo Li et al.
Summary: This article reviews ship detection and classification methods based on optical remote sensing images, analyzes feature extraction strategies and algorithms, summarizes publicly available datasets as benchmarks for verification, and provides insight into future development trends.
CHINESE JOURNAL OF AERONAUTICS
(2021)
Article
Engineering, Marine
Miaohui Zhang et al.
Summary: This paper presents an automatic ship detection approach in SAR images using phase spectrum, involving sea-land segmentation and ship detection based on phase spectrum. The proposed method is validated to be efficient through experimental results.
JOURNAL OF OCEAN ENGINEERING AND SCIENCE
(2021)
Article
Environmental Sciences
Bogdan Iancu et al.
Summary: The availability of domain-specific datasets is crucial in object detection, yet there is a limited number of studies on maritime vessel detection. The authors collected a dataset of maritime vessel images, annotated them accurately, and evaluated four prevalent object detection algorithms. The experiments showed that Faster R-CNN with Inception-Resnet v2 outperforms the others in most cases, except in the large object category where EfficientDet excels.
Article
Multidisciplinary Sciences
Yang Jie et al.
Summary: Ship detection and tracking in inland waterways is crucial for navigation safety. This paper introduces improvements to the YOLOv3 detection algorithm and its application in Deep SORT tracking algorithm, enhancing performance in ship detection and tracking. The enhanced algorithm shows increased precision and frame rate, with Deep SORT proving superior in complex scenarios.
Article
Computer Science, Interdisciplinary Applications
Won-Jae Lee et al.
Summary: In this study, ship awareness was achieved through ship detection using YOLO(v3), constructing a virtual image dataset using Unity, calculating the position of the horizon in ship images, tracking target ships' positions and speeds, and proposing a deep learning model to determine ship headings. The proposed method for ship awareness based on camera images was successfully validated in various scenarios, demonstrating good detection, localization, and tracking performance.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2021)
Article
Instruments & Instrumentation
Lin Li et al.
Summary: In dense sun-glint scenes, an effective method for infrared ship detection involves utilizing time fluctuation features and space structure features to accurately describe the difference between ships and clutters, thereby achieving high precision and recall rates.
INFRARED PHYSICS & TECHNOLOGY
(2021)
Article
Engineering, Marine
Ryan Wen Liu et al.
Summary: The study proposes an enhanced convolutional neural network and a flexible data augmentation strategy for improving ship detection under various weather conditions. Experimental results demonstrate the superior performance of the proposed method in terms of detection accuracy, robustness, and efficiency compared to other existing methods.
Proceedings Paper
Computer Science, Information Systems
Zhikang Qin et al.
Summary: The improved ship target detection method based on the YOLOv3 algorithm enhances the accuracy and real-time performance by incorporating the CBAM attention mechanism, SPP module, GIoU loss function, and ASFF method. The method effectively improves the detection accuracy of ship targets on the sea surface and meets the requirements of real-time detection.
ICECC 2021: 4TH INTERNATIONAL CONFERENCE ON ELECTRONICS, COMMUNICATIONS AND CONTROL ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Li Liu et al.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2020)
Article
Computer Science, Artificial Intelligence
Xiongwei Wu et al.
Article
Engineering, Industrial
Zhijun Chen et al.
Proceedings Paper
Computer Science, Software Engineering
Yitong Zheng et al.
2020 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO (ICME)
(2020)
Article
Environmental Sciences
Ramona Pelich et al.
Article
Environmental Sciences
Maria Daniela Graziano et al.
Proceedings Paper
Automation & Control Systems
Bo Fu et al.
2019 IEEE INTERNATIONAL CONFERENCE ON MECHATRONICS AND AUTOMATION (ICMA)
(2019)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Lin He et al.
PROCEEDINGS OF THE THIRD INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE AND APPLICATION ENGINEERING (CSAE2019)
(2019)
Article
Transportation Science & Technology
R. Glenn Wright
TRANSNAV-INTERNATIONAL JOURNAL ON MARINE NAVIGATION AND SAFETY OF SEA TRANSPORTATION
(2019)
Article
Engineering, Electrical & Electronic
Jia-chun Zheng et al.
MICROSYSTEM TECHNOLOGIES-MICRO-AND NANOSYSTEMS-INFORMATION STORAGE AND PROCESSING SYSTEMS
(2018)
Article
Computer Science, Theory & Methods
Ligang Zhang et al.
ACM COMPUTING SURVEYS
(2018)
Article
Computer Science, Information Systems
Zhenfeng Shao et al.
IEEE TRANSACTIONS ON MULTIMEDIA
(2018)
Proceedings Paper
Computer Science, Information Systems
Jamyoung Koo et al.
26TH ACM SIGSPATIAL INTERNATIONAL CONFERENCE ON ADVANCES IN GEOGRAPHIC INFORMATION SYSTEMS (ACM SIGSPATIAL GIS 2018)
(2018)
Article
Engineering, Civil
Dilip K. Prasad et al.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Kaiming He et al.
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Erhan Gundogdu et al.
COMPUTER VISION - ACCV 2016, PT V
(2017)
Article
Computer Science, Artificial Intelligence
Ross Girshick et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2016)
Proceedings Paper
Computer Science, Artificial Intelligence
Ross Girshick
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
(2015)
Article
Engineering, Marine
Zhi Zhao et al.
JOURNAL OF NAVIGATION
(2014)
Article
Computer Science, Information Systems
Wu-Chih Hu et al.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2011)
Review
Computer Science, Artificial Intelligence
WM Hu et al.
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART C-APPLICATIONS AND REVIEWS
(2004)
