Underwater object detection has long been severely affected by factors such as illumination conditions, suspended particles, and turbidity, which lead to weakened object textures, color distortion, and obscure morphological features. Existing deep learning detection frameworks based on visual features still exhibit unstable performance in these scenarios. To improve model adaptability, this paper proposes Bidirectional weighted Concat with Efficient multi-scale Attention You Only Look Once (BCEA-YOLO), a detection framework designed to mitigate underwater feature degradation through coordinated multi-scale processing. It employs an efficient multi-scale attention mechanism to suppress background noise while utilizing a bidirectional weighted concatenator to refine the fusion of visual features. To better capture tiny and occluded objects, a four-scale detection head with a P2 detection branch is implemented to preserve high-resolution spatial cues. Experimental results on two public benchmark datasets (URPC2018 and URPC2020) demonstrate that the proposed method outperforms competing methods, achieving mean average precision values of 78.5% and 85.5%, respectively, at an intersection over union threshold of 0.5. These results indicate the effectiveness of the proposed method for underwater object detection. Furthermore, edge deployment tests on the NVIDIA Jetson platform validate its real-time inference efficiency, providing a practical visual perception solution for autonomous underwater vehicles.
| [1] |
Abid S, Abid N. Advancing a sustainable blue economy through innovation: a systematic review. Discov Sustain, 2025, 6: 1092
|
| [2] |
Afifah V, Erniwati S (2026) YOLOv8 for object detection: a comprehensive review of advances, techniques, and applications. Int J Adv Comput Inf 2(1):53–61. https://doi.org/10.71129/ijaci.v2i1.pp53-61
|
| [3] |
Anwar S, Li CY. Diving deeper into underwater image enhancement: a survey. Signal Proc-Image Commun, 2020, 89: 115978
|
| [4] |
Er MJ, Chen J, Zhang YN, Gao WX. Research challenges, recent advances, and popular datasets in deep learning-based underwater marine object detection: a review. Sensors, 2023, 23(4 1990
|
| [5] |
Feng JF, Jin T. CEH-YOLO: a composite enhanced YOLO-based model for underwater object detection. Ecol Inform, 2024, 82: 102758
|
| [6] |
Folkman L, Pitt KA, Stantic B. A data-centric framework for combating domain shift in underwater object detection with image enhancement. Appl Intell, 2025, 55(4 272
|
| [7] |
Hu K, Lu FY, Lu MX, Deng ZL, Liu YP. A marine object detection algorithm based on SSD and feature enhancement. Complexity, 2020
|
| [8] |
Hu K, Weng CH, Zhang YW, Jin JL, Xia QF. An overview of underwater vision enhancement: from traditional methods to recent deep learning. J Mar Sci Eng, 2022, 10(2 241
|
| [9] |
Jian MW, Yang N, Tao C, Zhi HX, Luo HJ. Underwater object detection and datasets: a survey. Intell Mar Technol Syst, 2024, 2 9
|
| [10] |
Li CY, Liu WX, Gong GQ, Ding XB, Zhong X. SU-YOLO: spiking neural network for efficient underwater object detection. Neurocomputing, 2025, 644 130310
|
| [11] |
Li QM, Shi HW. YOLO-GE: an attention fusion enhanced underwater object detection algorithm. J Mar Sci Eng, 2024, 12(101885
|
| [12] |
Lin YH, Zhang YY, Yang DS, Zhou YKP, Wang LT. DLFE-YOLO: an enhanced framework for underwater object detection based on YOLOv8. Intell Mar Technol Syst, 2025, 3 37
|
| [13] |
Liu YR, Kong MR, Liu DS, Xu JL, Duan QL (2026) Advancements in computer vision-based fish multi-object tracking: a comprehensive review. Arch Computat Methods Eng 33:7397–7424. https://doi.org/10.1007/s11831-026-10554-2
|
| [14] |
Lu ZX, Liao LC, Xie XG, Yuan H. SCoralDet: efficient real-time underwater soft coral detection with YOLO. Ecol Inform, 2025, 85 102937
|
| [15] |
Ouyang DL, He S, Zhang GZ, Luo MZ, Guo HY, Zhan J et al (2023) Efficient multi-scale attention module with cross-spatial learning. In: ICASSP 2023–2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, pp 1–5. https://doi.org/10.1109/ICASSP49357.2023.10096516
|
| [16] |
Rashid M, Wang JF, Ahmed F, Ahmed N, Mohsan SAH, Alabdulkreem E, et al.. YOLOv8n-PFA: a parallel fusion attention network for enhanced target detection in challenging environments. Front Mar Sci, 2026, 13 1762170
|
| [17] |
Rzhanov Y, Lowell K. Addressing once more the (im)possibility of color reconstruction in underwater images. J Imaging, 2024, 10(10): 247
|
| [18] |
Saleem A, Awad A, Paheding S, Lucas E, Havens TC, Esselman PC. Understanding the influence of image enhancement on underwater object detection: a quantitative and qualitative study. Remote Sens, 2025, 17(2 185
|
| [19] |
Shakya B, El-Gayar O, Kaabi J, Mamun Ahmed K. Small-object detection at the edge: a pareto-efficient benchmark of lightweight YOLO models on UAV and overhead datasets. IEEE Access, 2026, 14: 528-548
|
| [20] |
Shen X, Wang HB, Cui TX, Guo ZC, Fu XP. Multiple information perception-based attention in YOLO for underwater object detection. Vis Comput, 2024, 40: 1415-1438
|
| [21] |
Shen X, Yuan GL, Wang HB, Fu XP. Unsupervised clustering optimization-based efficient attention in YOLO for underwater object detection. Artif Intell Rev, 2025, 58 219
|
| [22] |
Song P, Li P, Dai L, Wang T, Chen Z. Boosting R-CNN: reweighting R-CNN samples by RPN’s error for underwater object detection. Neurocomputing, 2023, 530: 150-164
|
| [23] |
Tan MX, Pang RM, Le QV (2020) EfficientDet: scalable and efficient object detection. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, pp 10778–10787. https://doi.org/10.1109/CVPR42600.2020.01079
|
| [24] |
Teng BW, Zhao HJ. Underwater target recognition methods based on the framework of deep learning: a survey. Int J Adv Rob Syst, 2020, 17(6): 1-12
|
| [25] |
Tsai YS, Tsai CT, Huang JH (2025) Multi-scale detection of underwater objects using attention mechanisms and normalized Wasserstein distance loss. J Supercomput 81(6):785. https://doi.org/10.1007/s11227-025-07251-5
|
| [26] |
Varghese R, Sambath M (2024) YOLOv8: a novel object detection algorithm with enhanced performance and robustness. In: 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS). IEEE, pp 1–6. https://doi.org/10.1109/ADICS58448.2024.10533619
|
| [27] |
Wang H, Sun SX, Bai X, Wang J, Ren P (2023) A reinforcement learning paradigm of configuring visual enhancement for object detection in underwater scenes. IEEE J Ocean Eng 48(2):443–461. https://doi.org/10.1109/JOE.2022.3226202
|
| [28] |
Woo S, Park J, Lee JY, Kweon IS (2018) CBAM: convolutional block attention module. In: Ferrari V (ed) Computer vision–ECCV 2018. Springer, Cham, pp 3–19. https://doi.org/10.1007/978-3-030-01234-2_1
|
| [29] |
Xiao ZH, Li ZP, Li HH, Li MT, Liu XY, Kong YY. Multi-scale feature fusion enhancement for underwater object detection. Sensors, 2024, 24(22 7201
|
| [30] |
Xu SB, Zhang MH, Song W, Mei HB, He Q, Liotta A (2023) A systematic review and analysis of deep learning-based underwater object detection. Neurocomputing 527:204–232. https://doi.org/10.1016/j.neucom.2023.01.056
|
| [31] |
Xu T, Yang ZH, Ye Q, Bai Y, Huang MF, Wang F et al (2026) Deep learning enhanced object detection for unmanned surface vehicles across diverse sensor types: a state-of-the-art review. Ships Offshore Struct [Online early access]. https://doi.org/10.1080/17445302.2026.2616615
|
| [32] |
Yang YK, Wang J, Wang WQ (2026) An efficient lightweight network for underwater small object detection. IEEE Access 14:29781–29792. https://doi.org/10.1109/ACCESS.2026.3665167
|
| [33] |
Yang YY, Chen L, Zhang J, Long LC, Wang ZF. UGC-YOLO: underwater environment object detection based on YOLO with a global context block. J Ocean Univ China, 2023, 22(3): 665-674
|
| [34] |
Zeng LC, Sun B, Zhu DQ (2021) Underwater target detection based on faster R-CNN and adversarial occlusion network. Eng Appl Artif Intell 100:104190. https://doi.org/10.1016/j.engappai.2021.104190
|
| [35] |
Zhang C, Wu S, Huang BH, Zhao BC, Cui FQ, Li XK (2026a) FSD-Net: underwater object detection based on frequency and spatial domain feature enhancement. Front Artif Intell 9:1770342. https://doi.org/10.3389/frai.2026.1770342
|
| [36] |
Zhang Y, Qin T, Zhou YM (2026b) A YOLO-OC real-time small object detection in ocean scenes. Comput Vis Image Underst 264:104625. https://doi.org/10.1016/j.cviu.2025.104625
|
| [37] |
Zhang ZH, Qu Y, Wang T, Rao Y, Jiang D, Li SW, et al.. An improved YOLOv8n used for fish detection in natural water environments. Animals, 2024, 14(14 2022
|
| [38] |
Zhao XY, Wang Z, Deng ZC, Qin HD (2024a) G-Net: an efficient convolutional network for underwater object detection. J Mar Sci Eng 12(1):116. https://doi.org/10.3390/jmse12010116
|
| [39] |
Zhao YY, Sun FJ, Wu XW (2024b) FEB-YOLOv8: a multi-scale lightweight detection model for underwater object detection. PLoS One 19(9):e0311173. https://doi.org/10.1371/journal.pone.0311173
|
| [40] |
Zhao ZQ, Zheng P, Xu ST, Wu XD. Object detection with deep learning: a review. IEEE Trans Neural Netw Learn Syst, 2019, 30(11): 3212-3232
|
| [41] |
Zhou WH, Zeng JB, Li S, Zhang YX. UW-YOLO-Bio: a real-time lightweight detector for underwater biological perception with global and regional context awareness. J Mar Sci Eng, 2025, 13(11 2189
|
| [42] |
Zhu XK, Lyu SC, Wang X, Zhao Q (2021) TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops. IEEE, pp 2778–2788. https://doi.org/10.1109/ICCVW54120.2021.00312
|
Funding
Department of Education of Liaoning Province(LJ222410146057)
Rights & permissions
The Author(s)
PDF
