Pengembangan Model Deteksi Sampah Berbasis YOLOV8 Dan Evaluasi Performanya Dalam Sistem Monitoring Lingkungan Sungai
Abstract
Pencemaran sungai akibat sampah merupakan permasalahan lingkungan yang membutuhkan solusi berbasis teknologi. Penelitian ini bertujuan untuk mengembangkan model deteksi sampah di permukaan sungai menggunakan algoritma YOLOv8 serta mengevaluasi performanya dalam sistem monitoring lingkungan. Dataset citra sampah dilabeli menggunakan Roboflow dan dilatih menggunakan YOLOv8 di Google Colaboratory. Model diuji dengan parameter pelatihan sebanyak 50 epoch, ukuran citra 320 × 320 piksel, dan batch size 32. Hasil pelatihan menunjukkan nilai precision sebesar 0.894, recall 0.833, mAP50 sebesar 0.89, dan mAP50-95 sebesar 0.726. Evaluasi lanjutan melalui confusion matrix dan pengujian terhadap 20 citra acak menunjukkan model mampu mendeteksi objek sampah dengan akurasi dan stabilitas yang baik dalam berbagai kondisi citra. Dengan demikian, model ini dinilai layak untuk diimplementasikan sebagai bagian dari sistem monitoring lingkungan sungai secara otomatis dan real-time.
Downloads
References
Md. K. Hasan, A. Shahriar, and K. U. Jim, “Water pollution in Bangladesh and its impact on public health,” Heliyon, vol. 5, no. 8, p. e02145, Aug. 2019, doi: 10.1016/j.heliyon.2019.e02145.
M. S. U. Chowdury et al., “IoT based real-time river water quality monitoring system,” in Procedia Computer Science, Elsevier B.V., 2019, pp. 161–168. doi: 10.1016/j.procs.2019.08.025.
Z.-Q. Zhao, P. Zheng, S.-T. Xu, and X. Wu, “Object Detection With Deep Learning: A Review,” IEEE Trans Neural Netw Learn Syst, vol. 30, no. 11, pp. 3212–3232, 2019, doi: 10.1109/TNNLS.2018.2876865.
J. Terven and D. Cordova-Esparza, “A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS,” Feb. 2024, doi: 10.3390/make5040083.
Z. Wu et al., “Using YOLOv5 for garbage classification,” in 2021 4th International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2021, Institute of Electrical and Electronics Engineers Inc., Aug. 2021, pp. 35–38. doi: 10.1109/PRAI53619.2021.9550790.
J. Redmon and A. Farhadi, “YOLOv3: An incremental improvement,” arXiv preprint arXiv:1804.02767, 2018.
H. Zhang, H. Yu, and G. Wang, “YOLO-based Object Detection for Real-time Industrial Applications,” IEEE Access, vol. 10, pp. 56353–56364, 2022.
W. Liu, Y. Lu, and X. Sun, “An Enhanced YOLO-Based Algorithm for Real-Time Object Detection,” Journal of Intelligent & Fuzzy Systems, vol. 44, no. 1, pp. 1079–1089, 2023.
C. Hamid, T. Wulandari, and R. Tjandrakirana, “Cloud-based Machine Learning Platform: A Case Study on Google Colab,” in IOP Conference Series: Materials Science and Engineering, 2021, p. 22066.
M. H. Hamid, F. Rahmatillah, and A. Kurniawan, “Optimization of Machine Learning Training Using Google Colab,” in Journal of Physics: Conference Series, 2021, p. 12005.
F. S. Uche, “Dataset Reusability Technique for Building YOLO AI Models Dataset Reusability Technique for Building YOLO Models.” [Online]. Available: https://www.researchgate.net/publication/351100077
R. Jain, M. Varshney, and R. Kumar, “Roboflow: A Comprehensive Platform for Data Management in Object Detection Tasks,” International Journal of Innovative Technology and Exploring Engineering (IJITEE), vol. 11, no. 5, pp. 89–93, 2022.
R. Jain, “An Efficient Workflow for Object Detection Tasks using Roboflow,” IJITEE, vol. 11, pp. 89–93, 2022.
A. Amaral, L. Pires, and R. Gonçalves, “Visual Studio Code in Collaborative Software Development: An Empirical Study,” Journal of Software Engineering and Applications, vol. 15, no. 1, pp. 1–16, 2022.
A. Amaral and others, “Effectiveness of Visual Studio Code as a Development Environment in Machine Learning Projects,” Journal of Systems and Software Engineering, vol. 11, no. 3, pp. 201–212, 2022.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
