基于YOLO的鲜茶叶智能分选系统的研究

Research on Intelligent Sorting System of Fresh Tea Based on YOLO

本文以茶叶图像为研究对象,开发一套基于YOLO的鲜茶叶智能分选系统。该系统分为硬件和软件部分,能实现实时监控茶园、远程控制云台、采集图像。通过引入所提出的多尺度特征融合模块,增强网络模型的特征提取与特征融合能力,并均衡提高茶叶中嫩芽的识别准确率;使用注意力机制提高网络模型对嫩芽目标特征信息的关注度,弱化复杂背景的干扰信息,提高网络模型的识别准确率,以实现茶叶嫩芽的自动检测。

This paper takes tea images as the research object and develops an intelligent sorting system for fresh tea based on YOLO.The system is divided into hardware and software parts,which can achieve real-time monitoring of tea gardens,remote control of cloud platforms,and image collection.By introducing the proposed multi-scale feature fusion module,the feature extraction and fusion capabilities of the network model are enhanced,and the recognition accuracy of tender buds in tea is balanced.Using attention mechanism to increase the attention of network models to the target feature information of tender buds,weaken the interference information of complex backgrounds,improve the recognition accuracy of network models,and achieve automatic detection of tea tender buds.