Scholarlink

<p>The increasing need for intelligent traffic management and autonomous surveillance has fuelled strong interest in real-time traffic monitoring systems. Major issues in this area include precise vehicle detection, segmentation, traffic density estimation, and the detection of fake or unauthorised license plates. This paper introduces a new AI-based architecture that incorporates YOLOv11, a sophisticated object detection model, and Easy-OCR, a compact optical character recognition (OCR) engine, to classify and detect High-Security Registration Plates (HSRP) in real-time. The system is architected for reliable operation across various environmental conditions, including low light, nighttime, and adverse weather conditions such as fog, rain, and snow. With deep learning and computer vision methods, the system detects cars and parses license plate data accurately in heavy traffic and visually impoverished scenes. An experimental assessment on a specially created dataset obtained under diverse weather and illumination conditions yielded an average vehicle detection accuracy of 96.4%, a plate reading accuracy of 94.8%, and an average processing rate of 25 frames per second (FPS). The model also recorded an F1-score of 0.95, surpassing traditional approaches in accuracy and efficiency. These findings indicate the capability of the proposed hybrid model for reliable, real-time traffic management. The end-to-end pipeline, from vehicle detection to license plate recognition, shows promising potential for mass deployment in intelligent traffic management systems, automated compliance checking, and city surveillance systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Hybrid AI Model Combining YOLOv11 and Easy-OCR for Real-Time Detection of Vehicles and High-Security Registration Plates in Intelligent Traffic Systems

  • 期刊论文

  • 发表日期：

    2026年5月20日

查看全文
  • Spoorthi T,
  • Shanthi Pichandi Anandaraj

摘要

The increasing need for intelligent traffic management and autonomous surveillance has fuelled strong interest in real-time traffic monitoring systems. Major issues in this area include precise vehicle detection, segmentation, traffic density estimation, and the detection of fake or unauthorised license plates. This paper introduces a new AI-based architecture that incorporates YOLOv11, a sophisticated object detection model, and Easy-OCR, a compact optical character recognition (OCR) engine, to classify and detect High-Security Registration Plates (HSRP) in real-time. The system is architected for reliable operation across various environmental conditions, including low light, nighttime, and adverse weather conditions such as fog, rain, and snow. With deep learning and computer vision methods, the system detects cars and parses license plate data accurately in heavy traffic and visually impoverished scenes. An experimental assessment on a specially created dataset obtained under diverse weather and illumination conditions yielded an average vehicle detection accuracy of 96.4%, a plate reading accuracy of 94.8%, and an average processing rate of 25 frames per second (FPS). The model also recorded an F1-score of 0.95, surpassing traditional approaches in accuracy and efficiency. These findings indicate the capability of the proposed hybrid model for reliable, real-time traffic management. The end-to-end pipeline, from vehicle detection to license plate recognition, shows promising potential for mass deployment in intelligent traffic management systems, automated compliance checking, and city surveillance systems.