Underwater Object Detection in Marine Ranching Based on Improved YOLOv8
Journal of Marine Science and Engineering · 25.12.2023 · 12 (1) · 55
DOI: 10.3390/jmse12010055 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Marine Science and Engineering, online erschienen am 25.12.2023. Verfasst von Rong Jia und 5 weiteren Autor:innen. Die Arbeit wurde bisher 31-mal zitiert (Crossref) und nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Qilu University of Technology (Shandong Academy of Sciences) Pilot Project of Science, Education, and Industry Integration Major Innovation Special Project “Pr…, Pivotal Technologies for Ocean Intelligent Sensing and Information Processing Based on End-to-End Cloud Architecture, National Natural Science Foundation of China.
Abstract (Originalsprache)
The aquaculture of marine ranching is of great significance for scientific aquaculture and the practice of statistically grasping existing information on the types of living marine resources and their density. However, underwater environments are complex, and there are many small and overlapping targets for marine organisms, which seriously affects the performance of detectors. To overcome these issues, we attempted to improve the YOLOv8 detector. The InceptionNeXt block was used in the backbone to enhance the feature extraction capabilities of the network. Subsequently, a separate and enhanced attention module (SEAM) was added to the neck to enhance the detection of overlapping targets. Moreover, the normalized Wasserstein distance (NWD) loss was proportionally added to the original CIoU loss to improve the detection of small targets. Data augmentation methods were used to improve the dataset during training to enhance the robustness of the network. The experimental results showed that the improved YOLOv8 achieved the mAP of 84.5%, which was an improvement over the original YOLOv8 of approximately 6.2%. Meanwhile, there were no significant increases in the numbers of parameters and computations. This detector can be applied on platforms for seafloor observation experiments in the field of marine ranching to complete the task of real-time detection of marine organisms.
Förderung
- Qilu University of Technology (Shandong Academy of Sciences) Pilot Project of Science, Education, and Industry Integration Major Innovation Special Project “Pr…
- Pivotal Technologies for Ocean Intelligent Sensing and Information Processing Based on End-to-End Cloud Architecture
- National Natural Science Foundation of China
- Project Plan of Pilot Project of Integration of Science, Education and Industry of Qilu University of Technology (Shandong Academy of Sciences)
Zitieren
APA
Jia, R., Lv, B., Chen, J., Liu, H., Cao, L., & Liu, M. (2023). Underwater Object Detection in Marine Ranching Based on Improved YOLOv8. Journal of Marine Science and Engineering, 12(1), 55. https://doi.org/10.3390/jmse12010055
Vancouver
Jia R, Lv B, Chen J, Liu H, Cao L, Liu M. Underwater Object Detection in Marine Ranching Based on Improved YOLOv8. Journal of Marine Science and Engineering. 2023;12(1):55. doi:10.3390/jmse12010055
BibTeX
@article{Jia2023,
author = {Jia, Rong and Lv, Bin and Chen, Jie and Liu, Hailin and Cao, Lin and Liu, Min},
title = {Underwater Object Detection in Marine Ranching Based on Improved YOLOv8},
journal = {Journal of Marine Science and Engineering},
year = {2023},
volume = {12},
number = {1},
pages = {55},
doi = {10.3390/jmse12010055}
}
EUVIMED