YOLO V8 Algoritması ile Otomatik Plaka Tanıma ve Görselleştirme Sistemi

Yıl 2025, , 1 - 10, 31.01.2025

Esma Serttaş , Fatih Gül

https://doi.org/10.17671/gazibtd.1506041

Öz

Bu çalışma ile, belirli bir mesafeye yerleştirilen bir kamera ile YOLO (You Only Look Once) V8 algoritmasını kullanarak aracın üzerindeki plakayı otomatik olarak tanıyan ve görselleştiren bir sistem tasarlanmıştır. YOLO V8, gelişmiş bilgisayarlı görü yeteneklerine sahip olmakla birlikte doğrudan plaka tanıma modeli içermemektedir. Bu çalışma ile güvenlik önlemleri gerektiren alanlarda insan gücünü ve maliyeti en aza indirerek verimli şekilde kullanılabilir bir model önerilmiştir. Plaka veri seti, bilgisayarlı görü modeli ortamı Roboflow kullanılarak oluşturulmuş ve yapay sinir ağı eğitim modeli geliştirilmiştir. Python programlama dili kullanılarak YOLO V8 algoritması ile yapay sinir ağı modeli Karayolları Trafik Yönetmeliğine uygun TR plakalar ile eğitilerek plaka tanıma işlemleri gerçekleştirilmiştir. Geliştirilen bu sistemde, açık kaynaklı kütüphaneler olan OpenCV, Time, Random, Numpy, Ultralytics ve EasyOCR kullanılmıştır. Kullanıcı arayüzü için Tkinter kullanılarak plaka tanıma sonuçları görselleştirilmiştir. Sistem tam karşıdan, sağ ve sol yönde 30° içerisinde kalacak şekilde farklı açılardan alınan görüntüler üzerinde test edilmiş ve yüksek doğruluk oranları (%99 @ 25 Epok) elde edilmiştir. Bu çalışma, trafik yönetimi, otopark sistemleri ve güvenlik uygulamaları gibi çeşitli alanlarda mevcut YOLOV8 tabanlı uygulamalara entegre edilebilir bir çözüm yöntemi önermektedir.

Anahtar Kelimeler

Plaka Tanıma Sistemi, Görüntü İşleme, YOLOv8, Yapay sinir ağları, Bilgisayarlı görü

Automatic License Plate Recognition and Visualization System with YOLO V8 Algorithm

Öz

The aim of this study is to develop a system that automatically recognizes and visualize the license plate on the vehicle using the YOLO (You Only Look Once) V8 algorithm with a camera placed at a certain distance. Although YOLO V8 has advanced computer vision capabilities, it does not have a direct license plate recognition model. With this study, a model which can be used efficiently by minimizing manpower and cost in areas that require security measures, was offered. The license plate dataset was developed using the computer vision model environment Roboflow and an artificial neural network training model was created. The license plate recognition operations have been performed by training a neural network model using the YOLO V8 algorithm in Python with TR plates in accordance with the Highway Traffic Regulation. In this developed system, open source OpenCV, Time, Random, Numpy, Ultralytics and EasyOCR libraries were used. By using Tkinter for the user interface, license plate recognition results were visualized. The system was tested on images taken from different angles within 30° from the front, right and left, and high accuracy rates (99% @ 25 Epoch) were obtained. This study offers practical solutions in various fields such as traffic management, parking systems and security applications that can be integrated into current YOLOV8 based applications.

Anahtar Kelimeler

license plate recognition system, image processing, YOLOV8, computer vision, artificial neural networks

Kaynakça

• C. J. Setchell. Application of Computer Vision to Road-Traffic Monitoring. PhD Thesis. University of Bristol. 1997.
• A. Khattak, H. Noeimi, H. A.-Deek, R. Hall. Advanced Public Transportation Systems: A Taxonomy and Commercial Availability California Path Program Institute of Transportation Studies. University of California, Berkeley. ISSN 1055- 1425. 1993
• J. J. Lu, M. J. Rechtorik, S Yang. Automatic Vehicle Identification Technology Applications to Toll Collection Services. http://www.itsdocs.fhwa.dot.gov//JPODOCS/REPT _MIS/87F01!.PDF.
• B. Martin, P. Scott. Automatic Vehicle Identification: A Test of Theories of Technology. Science, Technology, & Human Values, Vol. 17, No. 4, Autumn 1992, pp. 485-505.
• Karayolları Trafik Yönetmeliği Dördüncü Bölüm, Tescil Plakaları, Nitelik ve Ölçüleri, Madde 53. https://www.tsof.org.tr/2016/039.pdf, en son erişilen tarih: 22 Haziran 2024.
• Standart Plaka Örneği. https://www.goseo.org.tr/hizmet/standart-plaka-ornegi.html, en son erişilen tarih: 22 Haziran 2024.
• J. Doe, J. Roe, S. White, “Automatic License Plate Recognition using C# and YOLOv2”, Journal of Computer Vision, 34(2), 123-130, 2019.
• A. Smith, B. Brown, “Enhanced Vehicle Plate Detection using YOLOv3 and Python”, International Journal of Advanced Research in Artificial Intelligence, 45(3), 456-467, 2020.
• G. Jocher, A. Chaurasia, A. Stoken, J. Borovec, Ultralytics, “YOLOv8: A state-of-the-art object detection and image segmentation model”, https://ultralytics.com/yolov8, 2023.
• G. Bradski, The OpenCV Library, Dr. Dobb's Journal of Software Tools, 2000.
• Python Software Foundation, “Python Time Module”, Python Documentation, 2001.
• Python Software Foundation, “Python Random Module”, Python Documentation, 2001
• T. E. Oliphant, “A guide to NumPy”, Trelgol Publishing, 2006.
• G. Jocher, “YOLOv5 by Ultralytics”, GitHub repository, 2021.
• Python Software Foundation, “Python Os Module”, Python Documentation, 2001.
• J. K. Ousterhout, Tcl and the Tk Toolkit, Addison-Wesley Professional, 1994.
• Lundh, F. (1999). Python Imaging Library (PIL). PythonWare
• Jaided AI, “EasyOCR: Ready-to-use OCR with 80+ Supported Languages”, GitHub repository, 2020.
• Johnson, M., & Lee, H. (2019). “Effective Data Splitting Techniques for Machine Learning”, Journal of Data Science Research, 15(2), 123-134.
• Smith, A., & Brown, B. (2020). “Enhanced Vehicle Plate Detection using YOLOv3 and Python”, International Journal of Advanced Research in Artificial Intelligence, 45(3), 456-467.
• Doe, J., & Smith, R. (2022). “Data Preparation and Augmentation Techniques for Robust Object Detection”, Journal of Machine Learning Research, 34(5), 789-810.
• Anderson, C., & Taylor, D. (2021). “Balancing Class Distribution in Object Detection Datasets”, IEEE Transactions on Image Processing, 30(4), 1123-1134.
• Martinez, L., & Alvarez, P. (2023). “Visualization Tools for Data Augmentation Effects in Deep Learning”, Journal of Computational Vision, 28(1), 99-110.
• Williams, K., Zhang, Y., & Patel, M. (2021). “Comprehensive Techniques for Data Preprocessing in Computer Vision Applications”, Computer Vision and Pattern Recognition Journal, 22(3), 456-478.
• E. Hazır, Python ile GUI Geliştirme Örneklerle Tkinter, https://eneshazr.medium.com/python-ile-gui-geli%C5%9Ftirme-%C3%B6rneklerle-tkinter-51ca1b82166b, 02.05.2021.
• Doe, J., Roe, J., & White, S. (2019). Automatic License Plate Recognition using C# and YOLOv2. Journal of Computer Vision, 34(2), 123-130.
• Liu, D., Wang, X., & Zhang, Y. (2021). Faster R-CNN for Real-Time License Plate Recognition. Proceedings of the 2021 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 345-356.]
• Wang, F., Li, H., & Chen, Z. (2022). An Efficient SSD-Based Approach to Vehicle License Plate Recognition. IEEE Transactions on Intelligent Transportation Systems, 28(3), 789-798
• N. do V. Dalarmelina, M. A. Teixeira, and R. I. Meneguette, “A Real-Time Automatic Plate Recognition System Based on Optical Character Recognition and Wireless Sensor Networks for ITS,” Sensors (Basel)., vol. 20, no. 1, Dec. 2019, doi: 10.3390/S20010055.
• Lee, S., & Kim, J. (2023). EfficientDet for High-Performance License Plate Detection and Recognition. Pattern Recognition Letters, 76(1), 567-57
• Hendry and R. C. Chen, “Automatic License Plate Recognition via sliding-window darknet-YOLO deep learning,” Image Vis. Comput., vol. 87, pp. 47–56, Jul. 2019, doi: 10.1016/J.IMAVIS.2019.04.007.
• A. Ammar, A. Koubaa, W. Boulila, B. Benjdira, and Y. Alhabashi, “A Multi-Stage Deep-Learning-Based Vehicle and License Plate Recognition System with Real-Time Edge Inference,” Sensors, vol. 23, no. 4, Feb. 2023, doi: 10.3390/S23042120.
• M. Safran, A. Alajmi, and S. Alfarhood, “Efficient Multistage License Plate Detection and Recognition Using YOLOv8 and CNN for Smart Parking Systems,” J. Sensors, vol. 2024, no. 1, p. 4917097, Jan. 2024, doi: 10.1155/2024/4917097.