Optimasi Yolov11 Melalui Hyperparameter Tuning dan Data  Augmentasi untuk Meningkatkan Akurasi Deteksi Kendaraan pada Kondisi Malam Hari: Yolov11 Optimization Through Hyperparameter Tuning and Data Augmentation to Improve Vehicle Detection Accuracy at Night

Imam Alfath Zulkarnain; Kusrini Kusrini

doi:10.57152/malcom.v5i4.2250

Authors

Imam Alfath Zulkarnain Universitas Amikom Yogyakarta
Kusrini Kusrini Universitas Amikom Yogyakarta

DOI:

https://doi.org/10.57152/malcom.v5i4.2250

Keywords:

Deteksi Kendaraan, Penyesuaian Hiperparameter, Visi Komputer, YOLOv11

Abstract

Deteksi kendaraan pada malam hari menghadapi tantangan signifikan akibat pencahayaan rendah, silau lampu depan, dan kontras objek yang terbatas. Akurasi deteksi yang rendah pada malam hari menjadi penghambat utama dalam pengembangan sistem transportasi cerdas (ITS) dan sistem pengawasan lalu lintas yang andal secara 24/7. Penelitian ini bertujuan mengoptimalkan YOLOv11 untuk meningkatkan akurasi deteksi kendaraan dalam kondisi tersebut. Optimasi dilakukan melalui penyesuaian hiperparameter, termasuk pengaturan laju pembelajaran (0.001), momentum (0.937), dan weight decay (0.0005), serta penerapan teknik augmentasi data seperti penyesuaian saturasi dan kecerahan, translasi, skala, flipping horizontal, mosaic, dan mixup. Model diuji dalam dua skenario: (1) data malam hari dan (2) gabungan data siang dan malam. Hasil penelitian menunjukkan bahwa YOLOv11 yang telah dioptimalkan mencapai precision 0.97, recall 0.92, dan mAP@0.5 sebesar 0.97 pada skenario malam hari, melampaui kinerja YOLOv8 dan YOLOv11 baseline. Pada skenario gabungan, model tetap unggul dengan precision 0.95, recall 0.95, dan mAP@0.5 sebesar 0.98. Temuan ini membuktikan bahwa kombinasi penyesuaian hiperparameter dan augmentasi adaptif efektif meningkatkan kinerja deteksi kendaraan pada malam hari tanpa menurunkan akurasi pada kondisi siang. Pendekatan ini menjanjikan untuk diaplikasikan dalam sistem pemantauan lalu lintas berbasis visi komputer yang memerlukan konsistensi performa tinggi baik di siang maupun malam hari.

Downloads

Download data is not yet available.

References

G. D. Deepak and S. K. Bhat, "Optimizing YOLOv4 Hyperparameters for Enhanced Vehicle Detection in Intelligent Transportation Systems," International Journal of Intelligent Transportation Systems Research, 2025, doi: 10.1007/s13177-025-00519-3.

M. Humyam, T. Ashfaq, N. Z. Jhanjhi, and M. K. Alsadun, "Traffic Management: Multi-Scale Vehicle Detection in Varying Weather Conditions Using YOLOv4 and Spatial Pyramid Pooling Network," Electronics (Switzerland), vol. 11, no. 17, Sep. 2022, doi: 10.3390/electronics11172748.

S.-W. Huang, C.-T. Lin, S.-P. Chen, Y.-Y. Wu, P.-H. Hsu, and S.-H. Lai, "AugGAN: Cross Domain Adaptation with GAN-based Data Augmentation."

Y. Hu, "Improved YOLOv8 algorithm for vehicle image target detection based on learning rate optimization strategy," Applied and Computational Engineering, vol. 75, pp. 248-253, Aug. 2024, doi: 10.54254/2755-2721/7520240550.

H. Kumdake, C. Ongün, and A. Temizel, "Generative Data Augmentation for Vehicle Detection in Aerial Images," Dec. 2020, [Online]. Available: http://arxiv.org/abs/2012.04902.

C. T. Lin, S. W. Huang, Y. Y. Wu, and S. H. Lai, "GAN-Based Day-to-Night Image Style Transfer for Nighttime Vehicle Detection," IEEE Transactions on Intelligent Transportation Systems, vol. 22, no. 27, pp. 951-963, Feb. 2021, doi: 10.1109/TITS.2019.2961679.

A. Murumi and F. Murumi, "Data augmentation: A comprehensive survey of modern approaches," Dec. 01, 2022, Elsevier B.V. doi: 10.1016/j.array.2022.100258.

V. Ostankovich, R. Yagfarov, M. Rassabin, and S. Gafurov, "Application of CycleGAN-based Augmentation for Autonomous Driving at Night," in 2020 International Conference Nonlinearity, Information and Robotics, NIR 2020, Institute of Electrical and Electronics Engineers Inc., Dec. 2020. doi: 10.1109/NIR50484.2020.9290218.

J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," May 2016, [Online]. Available: http://arxiv.org/abs/1506.02640.

Y. Said, Y. Alassaf, R. Ghodhbani, T. Saidani, and O. Ben Rhaiem, "Optimized Convolutional Neural Networks with Multi-Scale Pyramid Feature Integration for Efficient Traffic Light Detection in Intelligent Transportation Systems," Computers, Materials and Continua, vol. 82, no. 2, pp. 3005-3018, 2025, doi: 10.32604/cmc.2025.060928.

D. Sun, F. Dornaika, and N. Barrena, "HSMix: Hard and soft mixing data augmentation for medical image segmentation," Information Fusion, vol. 115, Mar. 2025, doi: 10.1016/j.inffus.2024.102741.

D. Q. Tran, A. Aboah, Y. Jeon, M. Shoman, M. Park, and S. Park, "Low-Light Image Enhancement Framework for Improved Object Detection in Fisheye Lens Datasets," Apr. 2024, [Online]. Available: http://arxiv.org/abs/2404.10078.

Q. Yu, H. Liu, and Q. Wu, "An Improved YOLO for Road and Vehicle Target Detection Model," Journal of ICT Standardization, vol. 11, no. 2, pp. 197-216, 2023, doi: 10.13052/jicts2245-800X.1125.

Z. Yang and L. S. C. Pun-Cheng, "Vehicle detection in intelligent transportation systems and its applications under varying environments: A review," Jan. 01, 2018, Elsevier Ltd. doi: 10.1016/j.imavis.2017.09.008.

H. Zhang, Y. Gong, F. Yao, and Q. Zhang, "Research on Real-Time Detection Algorithm for Pedestrian and Vehicle in Foggy Weather Based on Lightweight XM-YOLOVIT," IEEE Access, vol. 12, pp. 7864-7883, 2024, doi: 10.1109/ACCESS.2023.3344666.

Q. Zhang, W. Guo, and M. Lin, "LLD-YOLO: a multi-module network for robust vehicle detection in low-light conditions," Signal Image Video Process, vol. 19, Aug. 2025, doi: 10.1007/s11760-025-03858-6.

J. Redmon and A. Farhadi, "YOLOv3: An Incremental Improvement," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 42, no. 6, pp. 1328-1336, June 2020, [Online]. Available: https://arxiv.org/abs/1804.02767.

A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "YOLOv4: Optimal Speed and Accuracy of Object Detection," Computer Vision and Image Understanding, vol. 212, Sep. 2021, Art. no. 103250, https://arxiv.org/abs/2004.10934.

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," Pattern Recognition Letters, vol. 174, pp. 12-19, Oct. 2023, https://arxiv.org/abs/2207.02696.

J. Terven, D. M. Córdova-Esparza, and J.-A. Romero-González, "A Comprehensive Review of YOLO: From YOLOv1 to YOLOv8 and Beyond," Archives of Computational Methods in Engineering, vol. 30, no. 4, pp. 2113-2130, May 2023, https://arxiv.org/abs/2304.00501.

J. Cao, T. Zhang, L. Hou, and N. Nan, "An improved YOLOv8 algorithm for small object detection in autonomous driving," Journal of Real-Time Image Processing, vol. 21, no. 4, Jul. 2024, doi: 10.1007/s11554-024-01517-6.