基于YOLOv8的河流漂浮垃圾检测算法研究
Detection of Floating Garbage in Rivers Based on Improved YOLOv8
席凯凯 1狄巨星 1杨阳1
作者信息
- 1. 河北建筑工程学院,河北 张家口 075000
- 折叠
摘要
为了解决无人机航拍的河流漂浮垃圾检测的难题,文章提出了一种改进YOLOv8的目标检测算法,用于检测河流漂浮的小目标垃圾.在原有YOLOv8算法模型的基础上,添加了 SENet注意力机制,提高了小目标检测的表征能力,并进行了超参数的微调.改进后的模型相比原有的YOLOv8s算法,平均精度(mAP)提升了 4.4％,达到80％.因此,改进的YOLOv8s更适用于小目标检测.
Abstract
In order to solve the problem of river floating garbage detection based on UAV aerial photography,this paper proposes an improved target detection algorithm of YOLOv8 for the de-tection of small target garbage floating in rivers.On the basis of the original YOLOv8 algorithm model,the SENet attention mechanism is added to improve the characterization ability of small target detection,and the hyperparameters are fine-tun ed.Compared with the original YOLOv8s algorithm,the average accuracy(mAP)of the improved model is increased by 4.4％to 80％.Therefore,the improved YOLOv8s is more suitable for small target detection.
关键词
YOLOv8/SENet/目标检测/超参数Key words
YOLOv8/SENet/object detection/Hyperparameters引用本文复制引用
出版年
2024
NETL
NSTL
万方数据