OCR Engines for License Plate Recognition: A Comparative Study of Tesseract, EasyOCR, PaddleOCR, and TrOCR

Afifah Khaerani Aziz; Marzuki Pilliang; Diana Kuniawati

doi:10.30871/jaic.v10i4.13437

Authors

Afifah Khaerani Aziz https://unsaka.ac.id
Marzuki Pilliang Salakanagara University
Diana Kuniawati https://unsaka.ac.id

DOI:

https://doi.org/10.30871/jaic.v10i4.13437

Keywords:

Automatic License Plate Recognition, Deep Learning, Intelligent Transportation Systems, Optical Character Recognition, Vehicle

Abstract

Automatic License Plate Recognition (ALPR) is a critical component of Intelligent Transportation Systems (ITS); however, the Optical Character Recognition (OCR) stage remains a significant bottleneck when confronting diverse linguistic scripts, complex plate formats, and environmental degradations. This systematic literature review comparatively evaluates four primary OCR engines—Tesseract, EasyOCR, PaddleOCR, and TrOCR—to bridge the gap between constrained local optimizations and globally resilient applications. Adhering to the PRISMA 2020 guidelines, a comprehensive search across six major academic databases spanning January 2020 to May 2026 initially identified 647 records. Following rigorous screening processes, a final core corpus of 21 empirical studies was qualitatively synthesized to account for extreme cross-study hardware and dataset heterogeneity. The analysis reveals that no single engine is universally superior; efficacy is fundamentally dictated by their underlying neural architectures and contextual deployment parameters. Tesseract offers maximum computational efficiency but fails significantly on non-Latin and complex scripts due to legacy segmentation limits. EasyOCR provides an optimal accuracy-to-speed ratio, making it highly suitable for real-time edge device deployments. PaddleOCR excels in robust sequential decoding, delivering the highest exact-match rates required for high-stakes applications like automated tolling. Conversely, the Transformer-based TrOCR emerges as the definitive frontier for highly complex, irregularly spaced, and multilingual plates (e.g., Han, CIS region scripts), though its severe computational latency currently restricts it to cloud-based infrastructures. Furthermore, the synthesis establishes that dynamic, environment-aware preprocessing is a mandatory prerequisite to mitigate visual stressors such as motion blur and low illumination. This review provides a novel, context-driven deployment taxonomy, equipping researchers and developers with actionable guidelines to navigate the trade-offs between architectural accuracy, script diversity, and computational resource constraints. Finally, the review identifies unified Large Vision-Language Models (LVLMs) as the next-generation trajectory to resolve current multi-stage pipeline dependencies.

Downloads

Download data is not yet available.

Author Biographies

Afifah Khaerani Aziz , https://unsaka.ac.id

Faculty of Science and Technology

Marzuki Pilliang, Salakanagara University

Faculty of Science and Technology

Diana Kuniawati, https://unsaka.ac.id

Faculty of Law and Business

References

[1] M. Ashkanani, “A Self-Adaptive Traffic Signal System Integrating Real-Time Vehicle Detection and License Plate Recognition for Enhanced Traffic Management,” Inventions, vol. 10, p. 14, 2025, doi: 10.3390/inventions10010014.

[2] E. Julianto, “Automatic Plate Number Recognition (APNR) using Deep Learning with YOLOv8 and TrOCR,” 2026, doi: 10.1109/ISIBER68248.2026.11470508.

[3] B. Buleu, “A Deep Learning-Based System for Automatic License Plate Recognition Using YOLOv12 and PaddleOCR,” Applied Sciences, vol. 15, p. 7833, 2025, doi: 10.3390/app15147833.

[4] R. N. Suharto, “Comparative Evaluation of YOLO-based Models for Indonesian License Plate Detection and Recognition with EASY OCR,” Procedia Comput. Sci., vol. 269, pp. 943–952, 2025, doi: 10.1016/j.procs.2025.09.037.

[5] Z. Ke, “TrOCR-based single/double Chinese license plates recognition,” International Conference on Optics, Electronics, and Communication Engineering (OECE 2025), p. 88, 2025, doi: 10.1117/12.3090329.

[6] I. Saitov, “CIS Multilingual License Plate Detection and Recognition Based on Convolutional and Transformer Neural Networks,” Procedia Comput. Sci., vol. 229, pp. 149–157, 2023, doi: 10.1016/j.procs.2023.12.016.

[7] M. J. Nayeem, “An Efficient Approach to Recognize Bangla License Plate for Diverse-Quality Images,” 2025, doi: 10.1007/978-981-96-3420-0_13.

[8] V. R. Thenisius, “Evaluating License Plate Recognition in Indonesia: A Comparison of OCRs on YOLOv8-Detected Plates,” 2025, doi: 10.1109/ICIC68054.2025.11308159.

[9] D. R. Vedhaviyassh, “Comparative Analysis of EasyOCR and TesseractOCR for Automatic License Plate Recognition using Deep Learning Algorithm,” 2022, doi: 10.1109/ICECA55336.2022.10009215.

[10] P. P. Reddy, “License Plate Detection using YOLO v8 and Performance Evaluation of EasyOCR, PaddleOCR and Tesseract,” 2024, doi: 10.1109/ICCCNT61001.2024.10725878.

[11] N. Shabbir, “Comparative Performance and Resource Utilization Analysis of OCR Models for Number Plate Recognition on Raspberry Pi 4,” 2025, doi: 10.1109/ComTech65062.2025.11034455.

[12] G. V. Reddy, “A comprehensive review on helmet detection and number plate recognition approaches,” E3S Web of Conferences, vol. 507, p. 1075, 2024, doi: 10.1051/e3sconf/202450701075.

[13] F. Delgado, S. Garrido, and B. S. Bezerra, “Barriers to Visibility in Supply Chains: Challenges and Opportunities of Artificial Intelligence Driven by Industry 4.0 Technologies,” Sustainability, vol. 17, p. 2998, 2025, doi: https://doi.org/10.3390/su17072998.

[14] M. Pilliang, Munawar, M. A. Hadi, G. Firmansyah, and B. Tjahjono, “Predicting Risk Matrix in Software Development Projects using BERT and K-Means,” in 2022 9th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI), Jakarta, Indonesia: IEEE, Oct. 2022, pp. 137–142. doi: 10.23919/EECSI56542.2022.9946637.

[15] Z. Munn, S. Moola, D. Riitano, and K. Lisy, “The development of a critical appraisal tool for use in systematic reviews addressing questions of prevalence,” Int. J. Health Policy Manag., vol. 3, no. 3, pp. 123–128, 2014, doi: 10.15171/ijhpm.2014.71.

[16] H. Aljzaere, “German License Plate Recognition System Using the YoloV8 Model and EasyOCR,” 2024, doi: 10.1109/ICSJ62869.2024.10804754.

[17] N. R. Adytia and G. P. Kusuma, “Indonesian license plate detection and identification using deep learning,” International Journal of Emerging Technology and Advanced Engineering, vol. 11, no. 7, pp. 1–7, Jul. 2021, doi: 10.46338/ijetae0721_01.

[18] M. L. Tham, “IoT Based License Plate Recognition System Using Deep Learning and OpenVINO,” 2021 4th International Conference on Sensors, Signal and Image Processing, pp. 7–14, 2021, doi: 10.1145/3502814.3502816.

[19] R. Parvathi, “Automated Vehicle Number Plate Detection Using Tesseract and Paddleocr,” Advances in Computational Intelligence and Robotics, pp. 90–107, 2023, doi: 10.4018/978-1-6684-9189-8.ch007.

[20] M. J. Nayeem, “Performance Evaluation of Tesseract and EasyOCR to Recognize Bangla License Plate for Diverse Quality Image,” 2024, doi: 10.1109/ICCIT64611.2024.11021754.

[21] M. H. Tusar, “Real Time Bangla License Plate Recognition with Deep Learning Techniques,” 2022, doi: 10.1109/IICAIET55139.2022.9936764.

[22] N. M. I. Abdul-Sattar, “AN INTELLIGENT SYSTEM FOR IRAQI ARABIC LICENSE PLATE RECOGNITION USING YOLO AND MACHINE LEARNING,” Kufa Journal of Engineering, vol. 17, pp. 417–436, 2026, doi: 10.30572/2018/KJE/170225.

[23] H. U. Iyer, “Automatic Number Plate and Face Recognition System for Secure Gate Entry into Military Establishments,” 2024, doi: 10.1109/ICSSES62373.2024.10561368.

[24] S. Rani, “Recognition of High-Security Registration Plate for Indian Vehicles,” 2023, doi: 10.1109/AISP57993.2023.10134888.

[25] V. Phon, “Integrating YOLOv5 and EasyOCR for Robust Thai License Plate Recognition,” 2025, doi: 10.1109/ECTI-CON64996.2025.11101195.

[26] P. Netinant, “Evaluating Factors Shaping Real-Time Internet-of-Things-Based License Plate Recognition Using Single-Board Computer Technology,” Technologies (Basel)., vol. 12, p. 98, 2024, doi: 10.3390/technologies12070098.

[27] H. Lin, “An End-to-End Recognition Model for Multi-Line Characters of Ship Plates Based on TrOCR,” 2025, doi: 10.1109/HPCC67675.2025.00238.

[28] G. A. Hakasin, “Impact of image pre-processing on enhancing paddleOCR performance for automatic indonesia number plate recognition,” Procedia Comput. Sci., vol. 269, pp. 1299–1308, 2025, doi: 10.1016/j.procs.2025.09.071.

[29] F. Karadag, “Reading Between the Blurs: A Comparative Analysis of Motion Deblurring Methods for Vehicle License Plate OCR,” Procedia Comput. Sci., vol. 269, pp. 1712–1722, 2025, doi: 10.1016/j.procs.2025.09.114.

[30] B. Xiao et al., “Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks,” in 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 4818–4829. doi: 10.1109/CVPR52733.2024.00461.

[31] Y. Liu et al., “OCRBench: on the hidden mystery of OCR in large multimodal models,” Science China Information Sciences, vol. 67, no. 12, Dec. 2024, doi: 10.1007/s11432-024-4235-6.