温室ROS喷雾机器人系统设计  

Design of Greenhouse Spray Robot System Based on ROS

作　　者：赵文锋 戚颖雯 张霞 郭柏帆 ZHAO Wenfeng;QI Yingwen;ZHANG Xia;GUO Bofan(College of Electronic Engineering,South China Agricultural University,Guangzhou 510642,China;College of Engineering,South China Agricultural University,Guangzhou 510642,China)

机构地区：[1]华南农业大学电子工程学院,广东广州510642 [2]华南农业大学工程学院,广东广州510642

出　　处：《佳木斯大学学报（自然科学版）》2024年第3期80-83,124,共5页Journal of Jiamusi University：Natural Science Edition

基　　金：广东省农业科技创新十大主攻方向“揭榜挂帅”项目(2022SDZG03);广东省科技创新战略专项资金项目(pdjh2022b0078)。

摘　　要：目前,我国温室内的传统植保作业主要以人工为主,劳动强度大、效率低,且会对环境造成污染。为解决以上问题,设计了一种基于ROS的温室喷雾机器人。该机器人采用深度相机进行自主规划路径;通过双目相机与YOLO v5机器视觉算法相结合,对番茄进行识别;最后利用喷雾系统对番茄实现喷雾的目标。实验结果表明:该机器人具有较好的定位效果和较高的识别准确率,并能够对番茄进行精准喷洒。该方案设计的机器人前景广阔,可靠性高,具有较好的应用价值。Currently,traditional greenhouse pest control operations in China are mainly manual,which are labor-intensive and low in efficiency,and may also cause environmental pollution.To address these issues,a greenhouse spray robot based on ROS was designed.This robot uses a depth camera for autonomous path planning,combined with a binocular camera and YOLOv5 machine vision algorithm,it can recognize tomatoes.Finally,it uses the spraying system to achieve the goal of spraying tomatoes.Experimental results have shown that this robot has good positioning performance,high recognition accuracy,and can accurately spray tomatoes.This solution has broad prospects,high reliability,and good application value.

关 键 词：ROS 喷雾 机器视觉 农业机器人 

分 类 号：S491[农业科学—植物保护]