基于深度学习的鹰嘴桃病虫害监测技术研究  

Research on Pest and Disease Monitoring Technology for Yingzui Peach Based on Deep Learning

作　　者：甘玉婉 曾静 周永福[1] 李泽航 胡展羽 林焕伟 GAN Yuwan;ZENG Jing;ZHOU Yongfu;LI Zehang;HU Zhanyu;LIN Huanwei(College of Electronic and Information Engineering,Heyuan Polytechnic,Heyuan 517000,China)

机构地区：[1]河源职业技术学院电子与信息工程学院,广东河源517000

出　　处：《现代信息科技》2025年第1期35-39,共5页Modern Information Technology

基　　金：广东省重点科研项目(2024ZDZX4089);广东省高校青年创新人才项目(2024KQNCX207);广东省高校青年创新人才项目(2023KQNCX233)。

摘　　要：针对传统病害鉴定方法效率不高、操作复杂等的问题,提出了一种基于深度学习的鹰嘴桃病虫害识别算法。该算法基于ResNet50网络架构,运用图像识别技术与深度学习相结合的方式,构建了一套高效的病虫害监测系统,能够迅速且准确地识别各类病虫害。首先针对河源连平地区鹰嘴桃的14种常见病虫害,建立了专门的数据集;其次利用深度残差网络模型进行了训练;最后利用PyQt5开发了图形化用户界面,实现了病虫害的自动诊断。这一自动诊断系统不仅高效、省力,而且环保,符合智慧农业的发展趋势,为用户提供了精准的病虫害防治手段。Aiming at the problems of low efficiency and complex operation of traditional disease identification methods,the pest and disease identification algorithm of Yingzui peach based on Deep Learning is proposed.Based on the ResNet50 network architecture,the algorithm uses the method of combining image recognition technology and Deep Learning to establish a high efficiency pest and disease monitoring system,which can quickly and accurately identify various types of pests and diseases.Firstly,a specialized dataset is established for 14 common pests and diseases of Yingzui peach in Lianping area of Heyuan.Secondly,the Deep Residual Network Model is used for training.Finally,the Graphical User Interface is developed using PyQt5 to realize the automatic diagnosis of pest and disease.This automatic diagnosis system is not only efficient and labor-saving but also environmentally friendly.It conforms to the development trend of smart agriculture and provides users with accurate control methods of pest and disease.

关 键 词：图像识别技术 病虫害监测 深度学习 病虫害数据集 智慧农业 鹰嘴桃病虫害 

分 类 号：T391[一般工业技术]