Scholarlink

<p>With advancements in technology and the economy, intelligent underwater robots are beginning to replace traditional manual fishing methods. However, the complex underwater environment can lead to image degradation, low resolution, and insufficient target detection samples. These factors present significant challenges for target detection. To address these issues, we propose an improved YOLOv5s model. Given the problem of excessive invalid information in underwater images, this method integrates a Coordinate Attention mechanism into the feature extraction network, enhances the feature extraction capability of the network, reduces feature redundancy, and helps the model locate underwater organism targets more quickly and accurately. Additionally, the Bi-directional Feature Pyramid Network is used to replace the Path Aggregation Network structure, and the features of different scales are fused quickly. Finally, the Structured Intersection over Union loss function is introduced to replace the Complete IoU (CIoU) loss function as a new boundary frame loss function, which improves the convergence speed and accuracy of the model. Experimental results demonstrate that the improved YOLOv5s model is superior to the state-of-the-art models in terms of both detection accuracy and speed. Code is available at <a href="https://github.com/910319491/UnderwaterYOLOv5">https://github.com/910319491/UnderwaterYOLOv5</a>.</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

Low visibility underwater biological target detection based on the improved YOLOV5s

  • 期刊论文

  • 发表日期：

    2025年6月22日

查看全文
  • Huixian Lin,
  • Haidong Deng,
  • Hong Du,
  • Yaohong Liu,
  • Junhua Xu

摘要

With advancements in technology and the economy, intelligent underwater robots are beginning to replace traditional manual fishing methods. However, the complex underwater environment can lead to image degradation, low resolution, and insufficient target detection samples. These factors present significant challenges for target detection. To address these issues, we propose an improved YOLOv5s model. Given the problem of excessive invalid information in underwater images, this method integrates a Coordinate Attention mechanism into the feature extraction network, enhances the feature extraction capability of the network, reduces feature redundancy, and helps the model locate underwater organism targets more quickly and accurately. Additionally, the Bi-directional Feature Pyramid Network is used to replace the Path Aggregation Network structure, and the features of different scales are fused quickly. Finally, the Structured Intersection over Union loss function is introduced to replace the Complete IoU (CIoU) loss function as a new boundary frame loss function, which improves the convergence speed and accuracy of the model. Experimental results demonstrate that the improved YOLOv5s model is superior to the state-of-the-art models in terms of both detection accuracy and speed. Code is available at https://github.com/910319491/UnderwaterYOLOv5.