Scholarlink

This study introduces a robust architectural framework for a globally compatible Automatic License Plate Recognition (ALPR) system, designed to recognize vehicles across international borders, including GCC countries. The system supports plates from vehicles produced between 1980 and 2019, tackling challenges such as varied formats, multilingual scripts, and environmental conditions. Leveraging Rekor’s AI technologies (Rekor Scout™, CarCheck™ APIs), it achieves 97.8 \(\%\) recognition accuracy overall and 98.6 \(\%\) for Kuwait private plates. The web-based system uses HTML5, TypeScript, and React.js on the frontend, with a Node.js/Express.js backend and a PostgreSQL/PostGIS database for geospatial indexing. It incorporates a GPS-free tracking method that reconstructs vehicle movements using strategic camera placement and path correlation, attaining 92 \(\%\) tracking accuracy despite discontinuous camera coverage. A comprehensive analytics suite provides real-time traffic visualization, historical movement trends, and vehicle classification via 15 dashboards. The system’s layered, microservices-based architecture ensures secure, scalable performance, handling up to 200 video streams with sub-120ms latency. A 5-node cluster demonstrated 208.3 req/sec throughput, while an edge-cloud hybrid model achieved 65ms latency with 58.5 \(\%\) resource use. Innovations include a multilingual recognition pipeline, strong data encryption and access controls compliant with international regulations, and a fault-tolerant distributed framework with 99.95 \(\%\) uptime. Compared to existing solutions, the system shows marked improvements, potentially reducing vehicle-related crimes by 35 \(\%\) and emergency response times by 42 \(\%\) . This work lays the groundwork for ethical, next-generation surveillance systems aligned with smart city goals.

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

Deep Learning Based Real Time Traffic Monitoring and Vehicle Tracking

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Hussain Alsulaisel,
  • Ali Alhawaj,
  • Haider Kamal,
  • Yousef Alyahyouh,
  • Salem Hussain,
  • Yehia Kotb

摘要

This study introduces a robust architectural framework for a globally compatible Automatic License Plate Recognition (ALPR) system, designed to recognize vehicles across international borders, including GCC countries. The system supports plates from vehicles produced between 1980 and 2019, tackling challenges such as varied formats, multilingual scripts, and environmental conditions. Leveraging Rekor’s AI technologies (Rekor Scout™, CarCheck™ APIs), it achieves 97.8 \(\%\) recognition accuracy overall and 98.6 \(\%\) for Kuwait private plates. The web-based system uses HTML5, TypeScript, and React.js on the frontend, with a Node.js/Express.js backend and a PostgreSQL/PostGIS database for geospatial indexing. It incorporates a GPS-free tracking method that reconstructs vehicle movements using strategic camera placement and path correlation, attaining 92 \(\%\) tracking accuracy despite discontinuous camera coverage. A comprehensive analytics suite provides real-time traffic visualization, historical movement trends, and vehicle classification via 15 dashboards. The system’s layered, microservices-based architecture ensures secure, scalable performance, handling up to 200 video streams with sub-120ms latency. A 5-node cluster demonstrated 208.3 req/sec throughput, while an edge-cloud hybrid model achieved 65ms latency with 58.5 \(\%\) resource use. Innovations include a multilingual recognition pipeline, strong data encryption and access controls compliant with international regulations, and a fault-tolerant distributed framework with 99.95 \(\%\) uptime. Compared to existing solutions, the system shows marked improvements, potentially reducing vehicle-related crimes by 35 \(\%\) and emergency response times by 42 \(\%\) . This work lays the groundwork for ethical, next-generation surveillance systems aligned with smart city goals.