Optimizing YOLOv8 Architecture and Augmentation for Efficient License Plate Detection

Muryan Awaludin; Yoke Lucia R Rehatalanit

doi:10.22146/jnteti.v15i2.24886

Muryan Awaludin Department of Information Systems, Faculty of Computer Science and Design, Universitas Dirgantara Marsekal Suryadarma, Kota Jakarta Timur, DKI Jakarta 13610, Indonesia
Yoke Lucia R Rehatalanit Department of Information Systems, Faculty of Computer Science and Design, Universitas Dirgantara Marsekal Suryadarma, Kota Jakarta Timur, DKI Jakarta 13610, Indonesia

DOI: https://doi.org/10.22146/jnteti.v15i2.24886

Keywords: Automatic Number Plate Recognition, Convolutional Neural Network, Data Augmentation, Edge Computing, Real-Time Systems, YOLOv8

Abstract

Automatic Number Plate Recognition (ANPR) is crucial for intelligent transportation systems but often falters in real-world conditions due to environmental variations. This study constructed a robust and computationally efficient vehicle license plate detection system that achieved high accuracy under diverse real-world challenges and was deployable on resource-constrained edge hardware for real-time operation. The proposed holistic framework integrated three key components: (a) the creation of the Dynamic Vehicle License Plate Dataset (DVLPD) v1.0, containing 866 annotated images with variations in lighting, weather, and camera angles; (b) the implementation of a targeted data augmentation pipeline employing geometric and photometric transformations to enhance model robustness; and (3) the architectural optimization of a You Only Look Once (YOLO) version 8 or YOLOv8 model through pruning, quantization, and hyperparameter tuning specifically for edge deployment. The optimized model achieved a mean average precision (mAP) of 91% on the test set. When deployed on a Raspberry Pi 4 in a prototype parking system, it demonstrated practical viability with an inference latency of 0.4 seconds per frame and an error rate of 4.2%. The results validate that the integration of a diverse dataset, strategic augmentation, and model optimization can yield an accurate and efficient ANPR solution suitable for real-time edge applications. Future work will focus on expanding the dataset to include more extreme conditions for greater generalization.

References

A.H.H. Bashori, Y.D. Rosita, and R.M. Akbar, “Penerapan metode OCR dan colour feature extraction untuk identifikasi pelat nomor kendaraan,” Semin. Nas. Fak. Tek., vol. 2, no. 1, pp. 43–48, Sep. 2023, doi: 10.36815/semastek.v2i1.122.

S. Eivazi et al., “Improving real-time CNN-based pupil detection through domain-specific data augmentation,” in ETRA '19, Proc. 11th ACM Symp. Eye Track. Res. Appl., 2019, pp. 1–6, doi: 10.1145/3314111.3319914.

K. Kumawat, A. Jain, and N. Tiwari, “Relevance of automatic number plate recognition systems in vehicle theft detection †,” Eng. Proc., vol. 59, no. 1, pp. 1–11, Jan. 2023, doi: 10.3390/engproc2023059185.

K.W. Leong et al., “License plate detection using deep learning object detection models,” in 2023 9th Int. Conf. Comput. Commun. Eng. (ICCCE), 2023, pp. 377–382, doi: 10.1109/ICCCE58854.2023.10246050.

Lubna, N. Mufti, and S.A.A. Shah, “Automatic number plate recognition: A detailed survey of relevant algorithms,” Sensors, vol. 21, no. 9, pp. 1–35, May 2021, doi: 10.3390/s21093028.

Z. Mahmood et al., “Towards automatic license plate detection,” Sensors, vol. 22, no. 3, pp. 1–19, Feb. 2022, doi: 10.3390/s22031245.

K.D. Rusakov, “Automatic modular license plate recognition system using fast convolutional neural networks,” 2020 13th Int. Conf. Manag. Large-Scale Syst. Dev. MLSD, 2020, pp. 1–4, doi: 10.1109/MLSD49919.2020.9247817.

S.A. Kadam, “Survey of feature extraction techniques for number plate detection,” Int. J. Sci. Res. Eng. Manag. (IJSREM), vol. 8, no. 3, pp. 1–5, Mar. 2024, doi: 10.55041/IJSREM29811.

P. Pabba et al., “Automatic number plate detection using deep learning and OCR,” in 2024 IEEE 4th Int. Conf. ICT Bus. Ind. Gov. (ICTBIG), 2024, pp. 1–5, doi: 10.1109/ICTBIG64922.2024.10911385.

A. Mohanty, A. Balasundaram, and A. Shaik, “A comprehensive analysis of deep learning frameworks to mitigate the impact of varied lighting and weather conditions,” in 2022 3rd Int. Conf. Smart Technol. Comput. Elect. Electron. (ICSTCEE), 2022, pp. 11–17, doi: 10.1109/ICSTCEE56972.2022.10100084.

M.M.H. Shuvo, S.K. Islam, J. Cheng, and B.I. Morshed, “Efficient acceleration of deep learning inference on resource-constrained edge devices: A review,” in Proc. IEEE, 2023, pp. 42–91, doi: 10.1109/JPROC.2022.3226481.

D. Liu et al., “Bringing AI to edge: From deep learning’s perspective,” Neurocomputing, vol. 485, pp. 297–320, May 2022, doi: 10.1016/j.neucom.2021.04.141.

N. Cahyadi, S. Salsabila, S. Raniprima, and V. Monita, “Review of license plate recognition techniques with deep learning,” J. Teknol., vol. 14, no. 2, pp. 458–469, Dec. 2024, doi: 10.51132/TEKNOLOGIKA.V14I2.415.

U. Sirisha et al., “Statistical analysis of design aspects of various YOLO-based deep learning models for object detection,” Int. J. Comput. Intell. Syst., vol. 16, pp. 1–29, Aug. 2023, doi: 10.1007/s44196-023-00302-w.

O. Hmidani and E.M.I. Alaoui, “A comprehensive survey of the R-CNN family for object detection,” in 2022 5th Int. Conf. Adv. Commun. Technol. Netw. (CommNet), 2022 pp. 1–6, doi: 10.1109/COMMNET56067.2022.9993862.

S. Kang et al., “Object detection YOLO algorithms and their industrial applications: Overview and comparative analysis,” Electronics, vol. 14, no. 6, pp. 1–36, Mar. 2025, doi: 10.3390/ELECTRONICS14061104.

T. Gorak and C.K.K. Reddy, “Helmet and number plate detection using YOLOV5 and YOLOV8,” Int. J. Interpret. Enigma Eng., vol. 1, no. 4, pp. 14–22, Dec. 2024, doi: 10.62674/IJIEE.2024.V1I04.003.

T. Rothmeier, W. Huber, and A.C. Knoll, “Time to shine: Fine-tuning object detection models with synthetic adverse weather images,” in 2024 IEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), 2024, pp. 4435–4444, doi: 10.1109/WACV57701.2024.00439.

U. Saha, B. Saha, and M.A. Imran, “A weather-adaptive convolutional neural network framework for better license plate detection,” Sensors, vol. 24, no. 23, pp. 1–16, Dec. 2024, doi: 10.3390/S24237841.

H. Hussain, P.S. Tamizharasan, and P.K. Yadav, “LCRM: Layer-wise complexity reduction method for CNN model optimization on end devices,” IEEE Access, vol. 11, pp. 66838–66857, Jun. 2023, doi: 10.1109/ACCESS.2023.3290620.

F. Ye, J. Wang, and C. Liu, “Misshaped boundary classifier model for license plate detection in haze weather using entropy CNN,” Inf. Sci., vol. 690, pp. 1–24, Feb. 2025, doi: 10.1016/J.INS.2024.121530.