An Intelligent Real-Time Detection and Classification System for Sustainable Aquatic Ecosystem Monitoring in Tropical Waters
DOI:
https://doi.org/10.52958/iftk.v22i2.12806Keywords:
Object Detection, real time, deep learning, aquatioc ecosiste, Yolo V8Abstract
Sustainable monitoring of aquatic ecosystems in tropical waters requires effective, adaptive, and intelligent technological approaches. This work proposes a design and evaluation of an intelligent real-time system for detecting and classifying freshwater fish species using You Only Look Once version 8 (YOLOv8), a recent deep learning architecture. The dataset used consists of 13,018 images of fishes constituting seven primary species: Catfish, Piranha, Tilapia, Betta, Milkfish, Gourami, and Koi. The research process includes preprocessing images (resizing and augmentation) and training the model using transfer learning techniques to expedite convergence and enhance accuracy. The evaluation findings show that the created system attained a maximum classification accuracy of 100% on the testing dataset. The model was successfully able to recognize species with distinct morphological traits, but a minor decrease in accuracy was reported in classifying fish with inductive body shapes. Overall results substantiate that YOLOv8 has solid potential as an efficient and replicable artificial intelligence-based approach to assisting sustainable aquatic ecosystem monitoring in tropical waters.
References
M. Al-Hassan and X. Wang, “Applications of computer vision for aquatic biodiversity monitoring: A review,” Ecol Inform, vol. 68, p. 101543, 2022, doi : https://doi.org/10.1016/j.ecoinf.2022.101543
A. M. Gonzalez and D. Lopez, “Monitoring fish behavior and population using computer vision and AI technologies,” Mar Ecol Prog Ser, vol. 650, pp. 145–158, 2020, doi: https://doi.org/10.3354/meps13423
F. A. O, “The State of World Fisheries and Aquaculture 2024: Blue Transformation in action,” Food and Agriculture Organization of the United Nations, Rome, 2024. doi: https://doi.org/10.4060/cc9700en
J. Kolding, P. A. M. van Zwieten, A. Suman, F. Poulain, and S. Funge-Smith, “Traditional inland fishery monitoring: examples and recommendations for resource-constrained countries,” Fish Manag Ecol, vol. 26, no. 5, pp. 433–446, 2019, doi: https://doi.org/10.1111/fme.12339
H. Li and D. Xu, “Real-time aquatic organism detection based on YOLO models for ecosystem monitoring,” Environ Monit Assess, vol. 194, no. 3, p. 158, 2022, doi: https://doi.org/10.1007/s10661-022-09743-z
P. Agarwal and R. Singh, “Deep learning-based real-time fish detection and classification using YOLOv8,” Journal of Intelligent Systems and Applications, vol. 15, no. 2, pp. 112–124, 2023, doi: https://doi.org/10.1016/j.jisa.2023.02.005
J. Chen, Y. Li, and X. Zhang, “Underwater image recognition using deep convolutional neural networks,” Applied Ocean Research, vol. 109, p. 102518, 2021, doi: https://doi.org/10.1016/j.apor.2021.102518
M. Rahim and M. Abdullah, “Evaluating deep learning models for fish species detection under varying environmental conditions,” J Mar Sci Eng, vol. 11, no. 2, p. 340, 2023, doi: https://doi.org/10.3390/jmse11020340
D. Akkaynak and T. Treibitz, “A revised physical model for underwater imaging and its application to image restoration,” IEEE Trans Pattern Anal Mach Intell, vol. 41, no. 7, pp. 1735–1742, 2018, doi: https://doi.org/10.1109/TPAMI.2018.2831252
C. Y. Wang, A. Bochkovskiy, and H. Y. M. Liao, “YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors,” 2022. [Online]. Available: http://arxiv.org/abs/2207.02696
G. Jocher, A. Chaurasia, and J. Qiu, “YOLOv8: Next-generation object detection and segmentation,” 2023. [Online]. Available: https://github.com/ultralytics/ultralytics
L. Tzutomi, “LabelImg,” 2018. [Online]. Available: https://github.com/tzutomi/labelImg
C. Shorten and T. M. Khoshgoftaar, “A survey on Image Data Augmentation for Deep Learning,” J Big Data, vol. 6, no. 1, p. 60, 2019, doi: https://doi.org/10.1186/s40537-019-0197-0
M. S. Khan, M. M. Rahman, and M. A. Hossain, “Automatic fish species classification using transfer learning and CNN architectures,” Journal of Aquatic Research and Development, vol. 13, no. 4, pp. 241–253, 2021, doi: https://doi.org/10.1007/s10661-021-09088-1
T. Y. Lin et al., “Microsoft COCO: Common Objects in Context,” 2014, pp. 740–755. doi: https://doi.org/10.1007/978-3-319-10602-1_48
D. P. Kingma and J. Ba, “Adam: A Method for Stochastic Optimization,” 2014. [Online]. Available: http://arxiv.org/abs/1412.6980
L. Prechelt, Early Stopping - But When? Springer, 1998. doi: https://doi.org/10.1007/3-540-49430-8_3
A. Paszke et al., “PyTorch: An Imperative Style, High-Performance Deep Learning Library,” 2019, pp. 8024–8035. [Online]. Available: https://proceedings.neurips.cc/paper/2019/file/bdbca288fee7eb8f802c659bd8be948a-Paper.pdf
J. J. Padilla, C. Spampinato, S. L. Cor-tés, J. E. Luján-García, and E. Diaz-Ramirez, “A Survey on Evaluation Metrics for Object Detection,” 2020. [Online]. Available: https://arxiv.org/abs/2007.13871
M. G. Pradana, “YOLOv8-Based Microplastic Detection and Quantification in River Water Microscopic Images,” Journal of Applied Data Sciences, vol. 7, no. 2, pp. 1533–1544, May 2026, doi: https://doi.org/10.47738/jads.v7i2.1262
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Nancy Jeane Tuturoong, Jimmy Reagen Robot, Djuwita Aling
This work is licensed under a Creative Commons Attribution 4.0 International License.
KEBIJAKAN YANG DIAJUKAN UNTUK JURNAL YANG MENAWARKAN AKSES TERBUKA
Syarat yang harus dipenuhi oleh Penulis sebagai berikut:
- Penulis menyimpan hak cipta dan memberikan jurnal hak penerbitan pertama naskah secara simultan dengan lisensi di bawah Creative Commons Attribution License yang mengizinkan orang lain untuk berbagi pekerjaan dengan sebuah pernyataan kepenulisan pekerjaan dan penerbitan awal di jurnal ini.
- Penulis bisa memasukkan ke dalam penyusunan kontraktual tambahan terpisah untuk distribusi non ekslusif versi kaya terbitan jurnal (contoh: mempostingnya ke repositori institusional atau menerbitkannya dalam sebuah buku), dengan pengakuan penerbitan awalnya di jurnal ini.
- Penulis diizinkan dan didorong untuk mem-posting karya mereka online (contoh: di repositori institusional atau di website mereka) sebelum dan selama proses penyerahan, karena dapat mengarahkan ke pertukaran produktif, seperti halnya sitiran yang lebih awal dan lebih hebat dari karya yang diterbitkan. (Lihat Efek Akses Terbuka).
