深度卷积神经网络在番茄叶病害识别中的应用  

Application of Deep Convolutional Neural Network in Tomato Leaf Disease Recognition

作　　者：鲍义东 李琳 粟小娓 BAO Yidong;LI Ling;SU Xiaowei(China Aerospace Science and Industry Corporation Huinong Technology General Department,Guizhou Aerospace Smart Agriculture Co.,Ltd,Guiyang,China,550025)

机构地区：[1]中国航天科工集团慧农科技总体部/贵州航天智慧农业有限公司,贵阳550025

出　　处：《福建电脑》2025年第4期21-25,共5页Journal of Fujian Computer

基　　金：2023年贵州省大数据局专项项目“现代山地智慧农场综合管控决策平台”(2023年1月-2024年12月)资助。

摘　　要：为提升复杂环境下农作物叶片病害的识别准确率,本文构建了一种识别番茄病害的卷积神经网络识别模型,并开发了一款适用于多种小程序平台的番茄常见病害图像识别小程序。实验结果表明,模型通过自动特征提取实现了95%以上的识别准确率,可为智能农业病害识别提供有力支持。To improve the recognition accuracy of crop leaf diseases in complex environments,this paper constructs a convolutional neural network recognition model for tomato diseases and develops a tomato common disease image recognition mini program suitable for multiple mini program platforms.The experimental results show that the model achieves a recognition accuracy of over 95%through automatic feature extraction,which can provide strong support for intelligent agricultural disease recognition.

关 键 词：卷积神经网络 番茄叶病害 深度学习 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]