嫁接机作业质量监测试验研究  

Study on Monitoring Graft Quality for Grafting Machine

作　　者：段志勇 张青[1] 初麒 何涛 辜松[1] Duan Zhiyong;Zhang Qing;Chu Qi;He Tao;Gu Song(College of Engineering,South China Agricultural University,Guangzhou 510642,China;Guangzhou SKY Mechanical&Electrical Co,Ltd.,Guangzhou 510642,China)

机构地区：[1]华南农业大学工程学院,广州510642 [2]广州实凯机电科技有限公司,广州510642

出　　处：《农机化研究》2024年第9期198-202,共5页Journal of Agricultural Mechanization Research

基　　金：广东省现代农业产业共性关键技术研发创新团队建设项目(2022KJ131)。

摘　　要：嫁接作业质量监测是提高茄果类嫁接机作业成功率的重要技术手段。为此,采用机器视觉的方法对嫁接过程中番茄嫁接苗的倾斜角度、弯曲角度、有无接穗等信息进行检测,以此为基础来判断嫁接苗的作业质量,进而为嫁接机嫁接作业和生产质量管理提供信息。试验结果表明:针对嫁接苗倾斜角度、弯曲角度,机器视觉测量相对人工测量的最大角度误差不超过2.15°;针对试验所选180株嫁接苗判断合格与否,与人工判断对比,视觉判断成功率为92.2%,故提出的方法可以对番茄嫁接苗的特征信息进行有效识别。此检测方法还可以调整相关参数,用于其他贴接类嫁接苗的作业质量特征识别应用。Grafting quality monitoring is an important technical means to improve the success rate of eggplant grafting machine.In this paper,the machine vision method is used to detect the inclination angle,bending angle and scion of tomato grafted seedlings in the grafting process.Based on this,we can judge whether the grafted seedlings are qualified or not,and guide the grafting operation of the grafting machine.The test results show that the maximum angle error of machine vision measurement relative to manual measurement is not more than 2.15°for inclination angle and bending angle;according to the 180 grafted seedlings selected in the experiment,the success rate of visual judgment was 92.2%compared with manual judgment.The method proposed in this paper can effectively identify the feature information of tomato grafted seedlings,and adjust the relevant parameters,which can be used for the feature recognition of other grafted seedlings.

关 键 词：嫁接机 茄果类砧木苗 机器视觉 分类 

分 类 号：S223.1[农业科学—农业机械化工程]