基于YOLOv5s的林业害虫目标检测方法分析
Analysis of Forest Pest Target Detection Method Based on YOLOv5s
陈中垚1
作者信息
摘要
阐述一种基于YOLOv5s模型的林业害虫目标检测方法,采用五类林业害虫的数据集对模型进行训练.实验结果分析表明,该模型能够较为准确地检测以及识别分类出小目标、多目标以及常态下的五类害虫图像.在对于图像预处理方面,实验使用多种数据增强的方法,有效地提高模型的泛化能力以及鲁棒性,对于提出的林业害虫目标检测模型具有良好的检测效果.
Abstract
This paper describes a forestry pest target detection method based on the YOLOv5s model,which is trained on a dataset of five types of forestry pests.The analysis of experimental results shows that the model can accurately detect and classify five types of pest images,including small targets,multiple targets,and normal conditions.In terms of image preprocessing,various data augmentation methods were used in the experiment,which effectively improved the model's generalization ability and robustness,and had a good detection effect on the proposed forestry pest target detection model.
关键词
计算机技术/YOLOv5s/害虫识别/目标检测/深度学习Key words
computer technology/YOLOv5s/pest recognition/object detection/deep learning引用本文复制引用
出版年
2024
NETL
NSTL
万方数据