基于YOLOv5的常见蔬菜图像检测方法  

A common vegetable image detection method based on YOLOv5

作　　者：蔡东吟 曹玉华[1] Cai Dongyin;Cao Yuhua(School of Intelligent Manufacturing Engineering,Guangdong Baiyun University,Guangzhou 510450,China)

机构地区：[1]广东白云学院智能制造工程学院,广州510450

出　　处：《现代计算机》2024年第18期45-48,共4页Modern Computer

基　　金：广东省净菜保鲜包装装备工程研究中心项目(2019GCZX008)。

摘　　要：蔬菜的准确分类与检测对蔬菜的生产加工效率具有重要意义。为了提升蔬菜在生产加工过程中的效率,提出了基于YOLOv5的常见蔬菜图像检测方法。首先,通过LabelImg工具对常见的6类蔬菜图像进行位置信息和类别的标注,为YOLOv5的训练提供数据集。其次,搭建基于YOLOv5的目标检测网络对蔬菜图片进行特征提取,获取蔬菜图像检测网络。最后,在蔬菜测试集上对本文方法的性能进行验证。实验结果表明,本文方法可以有效实现常见蔬菜的图像检测任务。The accurate classification and detection of vegetables are of great significance to their production and processing efficiency.In order to improve the efficiency of vegetable production and processing,a common vegetable image detection method based on YOLOv5 is proposed.First,,LabelImg tool was used to annotate the position information and category of 6 common vegetable images,so as to provide data set for YOLOv5 training,providing data set for YOLOv5 training.Secondly,a target detection network based on YOLOv5 is built to extract features from vegetable images and obtain the vegetable image detection network.Finally,the performance of the proposed method is verified on a vegetable test set.The experimental results show that the proposed method can effectively realize the image detection of common vegetables.

关 键 词：蔬菜 图像检测 目标检测 YOLOv5 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程] TS255.7[轻工技术与工程—农产品加工及贮藏工程]