Ship Detection and Tracking in Inland Waterways Using Improved YOLOv3 and Deep SORT
Open accessThe DOI matches its Crossref record; 37 of the 37 references it cites were found in a registry or an index.
Abstract
Ship detection and tracking is an important task in video surveillance in inland waterways. However, ships in inland navigation are faced with accidents such as collisions. For collision avoidance, we should strengthen the monitoring of navigation and the robustness of the entire system. Hence, this paper presents ship detection and tracking of ships using the improved You Only Look Once version 3 (YOLOv3) detection algorithm and Deep Simple Online and Real-time Tracking (Deep SORT) tracking algorithm. Three improvements are made to the YOLOv3 target detection algorithm. Firstly, the Kmeans clustering algorithm is used to optimize the initial value of the anchor frame to make it more suitable for ship application scenarios. Secondly, the output classifier is modified to a single Softmax classifier to suit our ship dataset which has three ship categories and mutual exclusion. Finally, Soft Non-Maximum Suppression (Soft-NMS) is introduced to solve the deficiencies of the Non-Maximum Suppression (NMS) algorithm when screening candidate frames. Results showed the mean Average Precision (mAP) and Frame Per Second (FPS) of the improved algorithm are increased by about 5% and 2, respectively, compared with the existing YOLOv3 detecting Algorithm. Then the improved YOLOv3 is applied in Deep Sort and the performance result of Deep Sort showed that, it has greater performance in complex scenes, and is robust to interference such as occlusion and camera movement, compared to state of art algorithms such as KCF, MIL, MOSSE, TLD, and Median Flow. With this improvement, it will help in the safety of inland navigation and protection from collisions and accidents.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures









Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Deep SORT uses spatial and temporal features in videos for locating the exact object tracking and it maintains the object IDs [47].”
From Visual Detection of Traffic Incident through Automatic Monitoring of Vehicle Activities (2024) · cites this paper as [47]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 9×
“It includes RCNN [10], Fast R-CNN [11], and Faster R-CNN [12].”
BackgroundMethodInfluentialDistinctive Image Features from Scale-Invariant Keypoints
Lowe, D. G. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000029664.99615.94 · Cited by 55,589 · cited 4×
“Fusion features based on HOG, SIFT (Scale-Invariant Feature Transform) and gray histogram features were extracted for ships.”
BackgroundInfluentialYou Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803 · cited 4×
“It includes YOLO [13], YOLOv2 [14], YOLOv3 [15], and SSD [16].”
BackgroundMethodInfluentialHistograms of Oriented Gradients for Human Detection
Dalal, N., & Triggs, B. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.177 · Cited by 33,584 · cited 2×
“Fusion features based on HOG, SIFT (Scale-Invariant Feature Transform) and gray histogram features were extracted for ships.”
BackgroundRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
A New Approach to Linear Filtering and Prediction Problems
Kalman, R. E. (1960) · Journal of Basic Engineering · doi:10.1115/1.3662552 · Cited by 31,751
(2016) · Cited by 31,590
Feature Pyramid Networks for Object Detection
Lin, T.-Y., Dollar, P., Girshick, R., He, K., Hariharan, B., & Belongie, S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.106 · Cited by 30,939 · cited 3×
“Multi-scale prediction For multi-scale prediction,YOLOv3 draws on the idea of the FPN algorithm [29] to predict in three scales, that is, the detection layer is used to detect on three feature maps of different sizes, and the step size is 32, 16, and 8.”
MethodInfluentialGirshick, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.169 · Cited by 28,105
SSD: Single Shot MultiBox Detector
Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.-Y., & Berg, A. C. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46448-0_2 · Cited by 22,128 · cited 3×
“It includes YOLO [13], YOLOv2 [14], YOLOv3 [15], and SSD [16].”
BackgroundMethod
Research works printed but not linked: 2
- Du, C.-J.; He, H.-J.; Sun, D.-W. Chapter 4-Object Classification Methods. In Computer Vision Technology for Food Quality Evaluation, 2nd ed.; Sun, D.-W., Ed.; Academic Press: San Diego, CA, USA, 2016; pp. 87-110, ISBN 978-0-12-802232-0.
- De Paz, J.F., Julián, V., Villarrubia, G., Marreiros, G., Novais, P., Eds.; Springer International Publishing: Cham, Switzerland, 2017; pp. 9-17.
Cited by
Rekavandi, A. M., Xu, L., Boussaïd, F., Seghouane, A., Hoefs, S., & Bennamoun, M. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3530678 · Cited by 161
“The components of the YOLOv3 network have been improved by Jie et al. [216] to achieve higher precision and recall values.”
An Image-Based Benchmark Dataset and a Novel Object Detector for Water Surface Object Detection
Zhou, Z., Sun, J., Yu, J., Liu, K., Duan, J., Chen, L., & Chen, C. L. P. (2021) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2021.723336 · Cited by 100
“Moreover, Jie et al. (2021) improved yolov3 to detect ships in inland waterways, the mAP and FPS of the improved method increased by about 5 and 2%.”
BackgroundMethodLiu, T., Pang, B., Zhang, L., Yang, W., & Sun, X. (2021) · Journal of Marine Science and Engineering · doi:10.3390/jmse9070753 · Cited by 92
“[38] modified the network classifier based on YOLO v3, and introduced the Soft-NMS algorithm.”
MethodDeepSORT: Real Time & Multi-Object Detection and Tracking with YOLO and TensorFlow
Pujara, A., & Bhamare, M. (2022) · International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) · doi:10.1109/icaiss55157.2022.10011018 · Cited by 59
“1: Flow Diagram of Basic steps in Object Tracking [22].”
BackgroundShipYOLO: An Enhanced Model for Ship Detection
Han, X., Zhao, L., Ning, Y., & Hu, J. (2021) · Journal of Advanced Transportation · doi:10.1155/2021/1060182 · Cited by 67
Bakırcı, M. (2024) · Regional Studies in Marine Science · doi:10.1016/j.rsma.2024.103975 · Cited by 78
A Lightweight Model for Real-Time Monitoring of Ships
Xing, B., Wang, W., Qian, J., Pan, C., & Le, Q. (2023) · Electronics · doi:10.3390/electronics12183804 · Cited by 37
Achieving Adaptive Visual Multi-Object Tracking with Unscented Kalman Filter
Zhang, G., Yin, J., Deng, P., Sun, Y., Zhou, L., & Zhang, K. (2022) · Sensors · doi:10.3390/s22239106 · Cited by 32
“[49] introduced the use of improved YOLOv3 and DeepSORT tracking algorithms to detect and track ships.”
MethodObject Recognition and Tracking in Moving Videos for Maritime Autonomous Surface Ships
Park, H., Ham, S.-H., Kim, T., & An, D. (2022) · Journal of Marine Science and Engineering · doi:10.3390/jmse10070841 · Cited by 29
Younes, O., Conruyt, N., Kayal, M., Martin, P. A., Lionel, B., Vignes‐Lebbe, R., Moussanif, H., & Zahir, J. (2025) · Ecological Informatics · doi:10.1016/j.ecoinf.2025.103170 · Cited by 32
“Improvements in ship detection and tracking have been achieved through enhanced YOLOv3 and DeepSORT algorithms, demonstrating advancements in target detection capabilities (Jie et al., 2021).”
Dong, K., Liu, T., Zheng, Y., Shi, Z., Du, H., & Wang, X. (2023) · Journal of Intelligent & Robotic Systems · doi:10.1007/s10846-023-02020-z · Cited by 25
“Jie et al. [44] introduced the Soft-NMS algorithm and also mod-ified the classifier of YOLO v3.”
Visual Detection of Traffic Incident through Automatic Monitoring of Vehicle Activities
Karim, A., Raza, M. A., Alharthi, Y. Z., Abbas, G., Othmen, S., Hossain, M. S., Nahar, A., & Mercorelli, P. (2024) · World Electric Vehicle Journal · doi:10.3390/wevj15090382 · Cited by 23
Li, A., Zhu, X., He, S., & Xia, J. (2021) · EURASIP Journal on Advances in Signal Processing · doi:10.1186/s13634-021-00831-6 · Cited by 20
“…introduced a residual network [22] Designed an adaptive feature fusion module and new loss function [23] Fused DenseNet in YOLOV3 [24] Improved YOLOv3 and Real-time tracking [25] High real-time Low accuracy Improve SSD [28] Lightweight feature optimizing network and feature fusion module [28] High”
BackgroundMethodReal-Time Fire Detection Method for Electric Vehicle Charging Stations Based on Machine Vision
Zhang, S., Yang, Q., Gao, Y., & Gao, D. (2022) · World Electric Vehicle Journal · doi:10.3390/wevj13020023 · Cited by 20
“The K-means algorithm is a division-based, circular, iterative clustering algorithm that uses distance to measure similarity between data objects [27], i.”
MethodWang, F., Yu, D., Huang, L., Zhang, Y., Chen, Y., & Wang, Z. (2024) · Geo-spatial Information Science · doi:10.1080/10095020.2024.2331552 · Cited by 16
“CNNs have been extensively employed and refined within the domain of remote sensing (Jie et al. 2021 The study of ship identification, instance segmentation, and detection utilizing Synthetic Aperture Radar (SAR) imagery is increasingly garnering substantial research interest.”
A Ship Tracking and Speed Extraction Framework in Hazy Weather Based on Deep Learning
Zhou, Z., Zhao, J., Chen, X., & Chen, Y. (2023) · Journal of Marine Science and Engineering · doi:10.3390/jmse11071353 · Cited by 13
“IOU matching and cascade matching are then performed between the predicted value and the detection frame information to track the trajectory of the target ship between the front and back frames of the video [36].”
MethodDong, K., Liu, T., Shi, Z., & Zhang, Y. (2024) · Journal of Real-Time Image Processing · doi:10.1007/s11554-024-01417-9 · Cited by 10
An analysis of research trends of inland waterway utilizing social network analysis
Choi, Y.-S., & Yeo, G. (2023) · The Asian Journal of Shipping and Logistics · doi:10.1016/j.ajsl.2023.02.002 · Cited by 9
Zhou, Z., Li, Z., Sun, J., Xu, L., & Zhou, X. (2023) · Journal of Marine Science and Engineering · doi:10.3390/jmse11081485 · Cited by 10
Video-Based Identification and Prediction Techniques for Stable Vessel Trajectories in Bridge Areas
Luo, W., Xia, Y., & He, T. (2024) · Sensors · doi:10.3390/s24020372 · Cited by 9
“To address this, Yang [18] optimized the anchor box initialization in the YOLOv3 algorithm and combined it with the DeepSORT [19] cascade tracking algorithm to achieve better results in vessel multi-object tracking.”
MethodYang, Y., Lee, S., & Lee, J. (2022) · Future Generation Computer Systems · doi:10.1016/j.future.2022.03.013 · Cited by 10
Real-Time Maritime Obstacle Detection Based on YOLOv5 for Autonomous Berthing
Chen, G., Qi, J., & Dai, Z. (2022) · Communications in computer and information science · doi:10.1007/978-981-19-1253-5_32 · Cited by 7
Toward Enhanced Support for Ship Sailing
Cafaro, M., Epicoco, I., Pulimeno, M., & Sansebastiano, E. (2023) · IEEE Access · doi:10.1109/access.2023.3303808 · Cited by 7
Moving Traffic Object Detection Based on Bayesian Theory Fusion
Sun, Y., Geng, K., Zhuang, W., Yin, G., Chen, X., Wang, J., & Ding, P. (2024) · Automotive Innovation · doi:10.1007/s42154-023-00245-0 · Cited by 5
“Jie et al. [40] and Gai et al. [41] adopted YOLO and Deep Sort to achieve multi-object tracking.”
Real-time tracking of recreational boats in coastal areas using deep learning
Signaroli, M., Lana, A., Cutolo, E., Alós, J., & González-Cid, Y. (2025) · Ocean & Coastal Management · doi:10.1016/j.ocecoaman.2025.107762 · Cited by 5
Manikandan, D. L., Arun, G., Pushpamalar, P., Dash, S. K., Usha, T., & Mallavarapu, V. R. (2025) · Marine Geophysical Research · doi:10.1007/s11001-025-09572-9 · Cited by 4
“The upgraded YOLOv3 detection algorithm and DeepSORT tracking algorithm are used in the model for ship identification and tracking that was developed by Jie et al. (2021).”
Guo, Y., Tian, X., & Xiao, Y. (2023) · Journal of Electronic Imaging · doi:10.1117/1.jei.32.4.043013 · Cited by 4
“Furthermore, Jie et al.17 enhanced YOLOv3 to detect ships in inland waterways, resulting in 5% improvement in mean average precision (mAP) and 2% increase in FPS. Recently, ShipYolo by Han et al.18 was introduced to solve the problem of missing detection of small vessels.”
BackgroundZou, S., Liu, J., Zhang, X., Wu, Z., Liu, J., & Han, B. (2025) · Complex & Intelligent Systems · doi:10.1007/s40747-025-01918-5 · Cited by 3
Multi-modal backend fusion for intelligent no-stop line violation detection in sea-connected locks
Lu, X., Xu, S., & Wu, S. (2025) · Measurement · doi:10.1016/j.measurement.2025.120018 · Cited by 3
Pieron, Ł., Jędrzychowski, K., Iwicki, S., & Durajczyk, P. (2026) · Sustainability · doi:10.3390/su18083844 · Cited by 3
Ship Detection and Tracking Based on a Custom Aerial Dataset
Paiano, L., Calabrese, F., Cataldo, M., Sebastiani, L., & Leonardi, N. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-06433-3_36 · Cited by 3
YOLOX-based ship target detection for Shore-based monitoring
Liu, J., & Sun, W. (2022) · doi:10.1145/3556384.3556420 · Cited by 3
“They eventually improved the robustness of the overall monitoring system [13].”
BackgroundMa, F., Jiang, X., Chen, C., Sun, J., Yan, X., & Wang, J. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3554717 · Cited by 2
“In the field of ship detection, Yang et al. [15] proposed a propagation detection and tracking algorithm based on the K-Means clustering algorithm and Soft Non-Maximum Suppression (Soft-NMS), aiming to enhance the algorithm’s accuracy and robustness.”
A Multi-Object Tracking framework with feature backtracking under severe occlusion
Ma, J., Zheng, Y., Lyu, Y., Jia, C., & Su, Y. (2026) · Ocean Engineering · doi:10.1016/j.oceaneng.2026.124175 · Cited by 4
Improve Detection and Tracking of Pedestrian Subclasses by Pre-Trained Models
Sukkar, M., Kumar, D., & Sindha, J. (2022) · Journal of Advanced Engineering and Computation · doi:10.55579/jaec.202263.369 · Cited by 2
“Pedestrians Tracking In a paper by Jie et al. 2021 [11] on Naval Ship Detection and Tracking, he proposed the developed Yolov3 as detection algorithm and the Deep-Sort as real-time deep tracking algorithm, through improvements they made to Yolo using the Kmeans clustering algorithm, then the output…”
MethodMaintaining Symmetry in Optimal and Safe Control of the Ship to Avoid Collisions at Sea
Lisowski, J. A. (2023) · Symmetry · doi:10.3390/sym15051016 · Cited by 2
Leveraging Multi-Object Tracking in Vision-based Target Following for Unmanned Aerial Vehicles
Ferreira, D. R., & Basiri, M. (2024) · IEEE International Conference on Autonomous Robot Systems and Competitions · doi:10.1109/icarsc61747.2024.10535936 · Cited by 2
“…for Unmanned Aerial Vehicles (UAV)s is an ever-evolving field of operations, essential in several areas that benefit from an airborne source of information such as: crowd management [2], ground [3] or oceanic [4] vehicle management, search and rescue missions [5] and auto-landing applications [6].”
BackgroundOptical resolution requirements for effective deep learning-based ship recognition
Zhang, Z., Liu, S., Xie, M., Li, Y., & Wang, H. (2023) · International Conference on Electromechanical Control Technology and Transportation (ICECTT) · doi:10.1117/12.2690113 · Cited by 1
A Ship Detection Method Based on Video Image in Inland Waterway
Li, Z., Zheng, J., Luo, Y., Wang, X., Wenjuan, E., Chen, Q., & Zhang, L. (2025) · International Conference on Control, Robotics and Intelligent System (CCRIS) · doi:10.1109/ccris65764.2025.11193052 · Cited by 1
“Jie et al. [5] improved the YOLOv3 detection algorithm and combined it with the Deep SORT tracking algorithm to enhance the accuracy of ship detection and tracking in inland waterways and the adaptability to complex scenarios.”
Dynamic path planning for bridge collision avoidance using trajectory clustering and BC-PPO fusion
Zhao, J., Tong, L., Peng, Z., Ma, W., Zeng, F., & Guo, Y. (2026) · Ships and Offshore Structures · doi:10.1080/17445302.2025.2607599 · Cited by 1
DeepTracking from Aerial Platforms
Sivakumar, A., Shahid, M., Muthukumar, S., & Balaji, A. (2021) · Asian Conference on Innovation in Technology (ASIANCON) · doi:10.1109/asiancon51346.2021.9544679 · Cited by 1
A Video Synopsis Algorithm Based on Target Detection for Ghosting Optimization
Liu, Y., Luo, L., Hong, W., Xin, Z., Chen, Y., & Yu, X. (2022) · International Conference on Electronic Information Technology and Computer Engineering (EITCE) · doi:10.1145/3573428.3573765 · Cited by 1
“In this paper, a target recognition algorithm based on deep learning neural network YOLO (you only look once) [16] and a target tracking algorithm based on Deep Sort (Simple Online And Real Time Tracking With A Deep Association Metric) [17] [18] are adopted [8] [9].”
MethodRETRACTED: Detection of an in-housed pig using modified YOLOv5 model
Devi, S. J., Doley, J., & Gupta, V. K. (2024) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-231032 · Cited by 1
RetractedLiu, Y., Luo, L., & Hong, W. (2023) · Frontiers in Computing and Intelligent Systems · doi:10.54097/fcis.v3i2.7702
An application of object detection in ship navigation
Dimitrelos, I. (2022) · DSpace - NTUA (National Technical University of Athens) · doi:10.26240/heal.ntua.23817
IR-YOLOv8 for lightweight object detection in inland waterway environments
Luo, J., Fu, Q., Wu, A., & Shen, J. (2026) · Scientific Reports · doi:10.1038/s41598-026-69839-y
Trajectory Tracking of Inland Ships Based on Improved Kalman Filter
Wang, Y., Deng, J., & Wang, J. (2026) · Journal of big data and computing. · doi:10.62517/jbdc.202601328
Metadata
- Authors
- Jie, Y., Leonidas, L., Mumtaz, F., & Ali, M.
- Affiliations
- Department of Information Engineering, Wuhan University of Technology, Wuhan 430070, China
- Quaid-i-Azam University (PK)
- Venue
- Symmetry
- Publisher
- MDPI AG
- Published
- 2021-02-12
- Volume
- 13
- Issue
- 2
- Pages
- 308
- ISSN
- 2073-8994 (ISSN-L)
- OpenAlex
- W3130098432
- Cited by
- 44
- References
- 37
- Citations by year
- 2026: 3 · 2025: 8 · 2024: 8 · 2023: 10 · 2022: 10 · 2021: 5
- FWCI
- 2.97
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Symmetry (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Key Research and Development Project of China (2020YFB1710800)
- National Nature Science Foundation of China (51879211)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Video Surveillance and Tracking Methods
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Keywords
- ship detection, vessel tracking, inland waterways, YOLOv3, DeepSORT, Soft-NMS, non-maximum suppression, collision avoidance, video surveillance, occlusion robustness, navigation safety
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.