Optimasi Arsitektur Deep Learning Berbasis YOLO untuk Deteksi Multi-Kelas Tingkat Kematangan Buah
Abstrak
Penilaian otomatis tingkat kematangan buah mangga merupakan tantangan dalam pertanian presisi karena perbedaan visual antar tingkat kematangan yang bersifat halus, dipengaruhi oleh variasi pencahayaan dan kondisi lingkungan alami. Pendekatan manual cenderung subjektif, sementara metode visi komputer konvensional masih kesulitan menangani kematangan bertingkat dan ketidakseimbangan data. Penelitian ini mengusulkan pengembangan model deteksi tingkat kematangan buah mangga berbasis YOLOv10s melalui studi ablasi sistematis. Model dikembangkan dengan memvariasikan mekanisme fusi fitur multi-skala, mekanisme attention, dan fungsi loss, serta dievaluasi pada dataset mangga dengan empat tingkat kematangan menggunakan metrik precision, recall, F1-score, dan mean Average Precision (mAP). Hasil eksperimen menunjukkan bahwa kombinasi BiFPN tiga lapisan, Hybrid Attention Transformer, dan fungsi loss EIoU memberikan performa terbaik dengan nilai [email protected] sebesar 0,96 serta keseimbangan precision dan recall yang baik. Hasil ini menunjukkan bahwa model yang diusulkan berpotensi diterapkan untuk deteksi tingkat kematangan buah mangga secara multi-tahap pada lingkungan pertanian nyata.
Kata kunci— attention mechanism, deteksi objek, feature fusion, kematangan buah mangga, YOLOv10s,
Referensi
I. Esaki, S. Noma, T. Ban, R. Sultana, and I. Shimizu, “Maturity Classification of Blueberry Fruit Using YOLO and Vision Transformer for Agricultural Assistance,” Sensors, vol. 23, no. 14, p. 6395, 2023.
R. Li, Y. Li, W. Qin, A. Abbas, S. Li, R. Ji, Y. Wu, Y. He, and J. Yang, “Lightweight Network for Corn Leaf Disease Identification Based on Improved YOLOv8s,” Computers and Electronics in Agriculture, vol. 217, p. 108687, 2024.
J. Tang, Z. Yu, and C. Shao, “Hybrid Attention Transformer Integrated YOLOv8 for Fruit Ripeness Detection,” Scientific Reports, vol. 15, p. 22652, 2025.
Y. Wang, C. Ouyang, H. Peng, J. Deng, L. Yang, H. Chen, Y. Luo, and P. Jiang, “YOLO-ALW: An Enhanced High-Precision Model for Chili Maturity Detection,” Agriculture, vol. 15, p. 1400, 2025.
J. J. K. Chai, J.-L. Xu, and C. O’Sullivan, “Real-Time Detection of Strawberry Ripeness Using Augmented Reality and Deep Learning,” Sensors, vol. 23, no. 17, p. 7639, 2023.
W. Huang, Y. Liao, P. Wang, Z. Chen, Z. Yang, L. Xu, and J. Mu, “AITP-YOLO: Improved Tomato Ripeness Detection Model Based on Multiple Strategies,” Frontiers in Plant Science, vol. 16, p. 1596739, 2025.
J. Ren, W. Wang, Y. Tian, and J. He, “YOLOv10-Kiwi: A YOLOv10-Based Lightweight Kiwifruit Detection Model in Trellised Orchards,” Frontiers in Plant Science, vol. 16, p. 1616165, 2025.
L. Wang, R. Ye, Y. Chen, and T. Li, “YOLOv10-LGDA: An Improved Algorithm for Defect Detection in Citrus Fruits Across Diverse Backgrounds,” Plants, vol. 14, p. 1990, 2025.
H. Fu, Z. Guo, Q. Feng, F. Xie, Y. Zuo, and T. Li, “MSOAR-YOLOv10: Multi-Scale Occluded Apple Detection for Enhanced Harvest Robotics,” Horticulturae, vol. 10, p. 1246, 2024.
P. Kamat, S. Gite, H. Chandekar, L. Dlima, and B. Pradhan, “Multi-class Fruit Ripeness Detection Using YOLO and SSD,” Journal of Food Engineering and Automation, vol. 9, no. 3, pp. 112--124, 2022.
A. A. Luthfie and A. Alamsyah, “Identifying Coconut Maturity Levels Using CNN and YOLOv8,” Agriculture, vol. 14, no. 4, p. 812, 2024.
Y. Xu, H. Li, Y. Zhou, Y. Zhai, Y. Yang, and D. Fu, “GLL-YOLO: A Lightweight Network for Detecting the Maturity of Blueberry Fruits,” Agriculture, vol. 15, p. 1877, 2025.
Z. Huang, X. Li, S. Fan, Y. Liu, H. Zou, X. He, S. Xu, J. Zhao, and W. Li, “ORD-YOLO: A Ripeness Recognition Method for Citrus Fruits in Complex Environments,” Agriculture, vol. 15, p. 1711, 2025.
B. Ma, Z. Hua, Y. Wen, H. Deng, Y. Zhao, L. Pu, and H. Song, “Using an Improved Lightweight YOLOv8 Model for Real-Time Detection of Multi-Stage Apple Fruit,” Artificial Intelligence in Agriculture, vol. 11, pp. 70--82, 2024.
C. Wang, H. Wang, Q. Han, Z. Zhang, D. Kong, and X. Zou, “Strawberry Detection and Ripeness Classification Using YOLOv8+,” Agriculture, vol. 14, p. 751, 2024.
F. Yuan, J. Wang, W. Ding, S. Mei, C. Fang, S. Chen, and H. Zhou, “A Lightweight and Rapid Dragon Fruit Detection Method,” Agriculture, vol. 15, p. 1120, 2025.
M. Nergiz, “Enhancing Strawberry Harvesting Efficiency through YOLO-v7 Object Detection Assessment,” Turkish Journal of Science and Technology, vol. 18, pp. 519--533, 2023.
H. C. Bazame, J. P. Molin, D. Althoff, and M. Martello, “Detection of Coffee Fruits on Tree Branches Using Computer Vision,” Scientia Agricola, vol. 80, p. e20220064, 2023.
R. E. Saragih, A. Purnajaya, I. Syafrinal, Y. Pernando, and Yodi, “Mango and Banana Ripeness Detection Based on Lightweight YOLOv8,” Jurnal Buana Informatika, vol. 15, no. 2, pp. 79--88, 2024.
A. Aljaafreh, E. Elzagzoug, J. Abukhait, A.-H. Soliman, S. Alja’afreh, A. Sivanathan, and J. Hughes, “A Real-Time Olive Fruit Detection for Harvesting Robot Based on YOLO Algorithms,” Journal of Robotics and Control, vol. 4, no. 2, pp. 111--121, 2023.
X. Gao, J. Du, X. Liu, D. Jia, and J. Wang, “Object Detection Based on Improved YOLOv10 for Electrical Equipment,” Processes, vol. 13, no. 529, 2025.
B.Liu, X.Wei, Q.Chen, J. Liu, Y. Chen, P. Yu, S. Lei, and Y. Hu, “Safety Helmet Detection Methods in Heavy Machinery Factory,” 2024.
F. Zhincheng, Y. Jie, C. Lifang, C. Zhichao, L. Linhong, “An Intelligent Waste-Sorting and Recycling Device Based on Improved EfficientNet”, International Journal of Environmental Research and Public Health (IJERPH), vol 19(23):15987.
X. Yang, J. Zhao, L. Zhao, I. Ganchev, et al., “Detection of River Floating Garbage Based on Improved YOLOv5,” Mathematics, vol. 10, no. 22, p. 4366, 2022, doi: 10.3390/math10224366.
H. T. Nguyen, T. Le, T. V. Nguyen, and N. T. Nguyen, “Enhancing MRI Brain Tumor Segmentation with an Additional Classification Network,” 2020, doi: 10.48550/arXiv.2009.12111.
Z. Wang, S. Wan, and X. Ma, “Remote Sensing Image Dense Target Detection Based on Rotating Frame,” Journal of Physics: Conference Series, vol. 2006, no. 1, p. 012049, 2021, doi: 10.1088/1742-6596/2006/1/012049.
