融合YOLOv5与ASFF算法的海产品目标检测算法研究

Research on Seafood Object Detection Algorithm Integrating YOLOv5 and ASFF Algorithm

针对水下目标检测识别精度低的问题,提出了一种基于改进YOLOv5的水下目标检测方法。通过对比多个注意力机制模块,在YOLOv5骨干网络引入了全局注意力机制模块,增强了特征提取,提高了采集特征的能力,并在YOLOv5模型上融合了自适应空间特征融合算法,实现底层特征与顶层特征融合。验证结果表明,所提算法的识别精度优于原始的YOLOv5算法,平均精度提升了8.5%,检测速度为76帧/秒。

To deal with the problem of low accuracy of underwater target detection and recognition,an underwater target detection method based on improved YOLOv5 is proposed.By comparing multiple attention mechanism modules,the global attention mechanism module is introduced in the YOLOv5 backbone network,which enhances feature extraction and improves the ability to collect features,and the adaptive spatial feature fusion algorithm is integrated on the YOLOv5 model to realize the fusion of underlying features and top-level features.The experimental results show that the recognition accuracy of the proposed algorithm is better than that of the original YOLOv5 algorithm.The average accuracy is improved by 8.5%,and the detection speed is 76 frames per second.