Scholarlink

The flourishing demand for next-gen parking solutions has accentuated the significance of accurate vehicle detection, and counting. To aid with streamlined detection and tracking of the vehicles, we have conducted this research work. In this work, we have experimented and analyzed various hybrid deep learning frameworks, to enhance real-time vehicle detection and counting. The work also includes a robust module for automatic number plate identification. The performance of these various models in the detection and tracking of vehicles from different vehicle footage and their comparative study is depicted in this work. Experimental results on the vehicular clips highlight that the proposed hybrid models outshine the conventional methods in terms of precision and computational efficiency. The best performing models were YOLOv8+Centroid Tracker+GRU and YOLOv11+Centroid Tracker+LSTM+GRU.

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

Enhanced Vehicle Detection and Tracking Using YOLO and RNN-Based Hybrid Models

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shriyans Rout,
  • Mahesh Jangid

摘要

The flourishing demand for next-gen parking solutions has accentuated the significance of accurate vehicle detection, and counting. To aid with streamlined detection and tracking of the vehicles, we have conducted this research work. In this work, we have experimented and analyzed various hybrid deep learning frameworks, to enhance real-time vehicle detection and counting. The work also includes a robust module for automatic number plate identification. The performance of these various models in the detection and tracking of vehicles from different vehicle footage and their comparative study is depicted in this work. Experimental results on the vehicular clips highlight that the proposed hybrid models outshine the conventional methods in terms of precision and computational efficiency. The best performing models were YOLOv8+Centroid Tracker+GRU and YOLOv11+Centroid Tracker+LSTM+GRU.