Scholarlink

<p>To address the challenge of balancing accuracy and real-time performance in ship detection for ports and waterways, this paper proposes an optimized detection algorithm named YOLOv11-FFCM, based on an enhanced YOLOv11 framework. The proposed method integrates the Focal Modulation module, FasterNetBlock structure, and CBAM attention mechanism to strengthen feature extraction and improve detection performance in scenarios with complex backgrounds, small targets, and low visibility. The experimental results show that the YOLOv11-FFCM achieves superior detection accuracy while maintaining an effective trade-off between computational complexity and real-time performance. These advantages make it highly suitable for practical applications in port and waterway ship surveillance. Compared with YOLOv11, YOLOv11-FFCM improves 4.4% of mAP@0.5 and 4.4% of mAP@0.5:0.95 on the HarbourShip dataset and 2.2% of mAP@0.5 and 2.4% of mAP@0.5:0.95 on the Unreal-Vessels-detection-v1 dataset.</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

Real-Time Detection Algorithm of Port Ships Based on Improved YOLOv11

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月17日

查看全文
  • Boweng Xing,
  • Qingsong Xing,
  • Chengwu Pan,
  • Qibo Le,
  • Fujia Bi

摘要

To address the challenge of balancing accuracy and real-time performance in ship detection for ports and waterways, this paper proposes an optimized detection algorithm named YOLOv11-FFCM, based on an enhanced YOLOv11 framework. The proposed method integrates the Focal Modulation module, FasterNetBlock structure, and CBAM attention mechanism to strengthen feature extraction and improve detection performance in scenarios with complex backgrounds, small targets, and low visibility. The experimental results show that the YOLOv11-FFCM achieves superior detection accuracy while maintaining an effective trade-off between computational complexity and real-time performance. These advantages make it highly suitable for practical applications in port and waterway ship surveillance. Compared with YOLOv11, YOLOv11-FFCM improves 4.4% of mAP@0.5 and 4.4% of mAP@0.5:0.95 on the HarbourShip dataset and 2.2% of mAP@0.5 and 2.4% of mAP@0.5:0.95 on the Unreal-Vessels-detection-v1 dataset.