基于物联网的跨区域果园病虫害监控预警仿真  

Simulation of Cross-Regional Orchard Pest Monitoring and Early Warning Based on Internet of Things

作　　者：徐艳玲 杨真[2] XU Yan-ling;YANG Zhen(Computer Science and Technology Department,Hanjiang Normal University,Shiyan Hubei 442700,China;Network&Information Center,East China Jiaotong University,Nanchang Jiangxi 330013,China)

机构地区：[1]汉江师范学院数学与计算机科学学院,湖北十堰442700 [2]华东交通大学网络信息中心,江西南昌330013

出　　处：《计算机仿真》2024年第6期583-587,共5页Computer Simulation

摘　　要：跨区域果园病虫害图像需要处理大量的数据,且果园病虫害图像的质量因环境、设备影响,会产生模糊、暗淡等现象,使得跨区域果园病虫害图像监控预警效率降低,提出基于物联网的跨区域果园病虫害图像监控预警。基于物联网技术建立跨区域果园病虫害图像监控预警平台;在物联网的网络层上通过暗原色优先算法和伽玛校正(Gamma correction, Gamma)方法,增强果园监控图像;利用Relief特征选择(Relief Feature Selection, Relief-F)算法实现果园监控图像的病虫害的优化特征提取;基于径向基函数(Radial Basis Function, RBF)神经网络方法完成跨区域果园病虫害图像识别,从而实现最终的跨区域果园病虫害图像监控预警。实验结果表明,所提方法的跨区域果园病虫害图像识别预警效果好,识别检测率高达97%、整体应用效果更好。Cross regional orchard pest and disease images require processing a large amount of data,and the quality of orchard pest and disease images may be blurred,dim,and other phenomena due to environmental and e⁃quipment factors,which reduces the efficiency of cross regional orchard pest and disease image monitoring and warn⁃ing.Therefore,a cross regional orchard pest and disease image monitoring and warning based on the Internet of Things is proposed.Firstly,a platform for monitoring and early warning was constructed by IoT technology.At the network layer of the IoT,the dark primary color priority algorithm and Gamma correction were utilized to enhance the monito⁃ring images.Subsequently,the Relief-F feature selection algorithm was employed to optimize and extract the features related to pests and diseases from orchard monitoring images.Furthermore,the Radial Basis Function(RBF)neural network method was adopted to complete the identification of pest and disease images in cross-regional orchard,thus achieving the final monitoring and early warning.The experimental results prove that the proposed method provides ef⁃fective monitoring and early warning for pest and disease images of cross-regional orchard.Meanwhile,the identifica⁃tion rate is up to 97%,and overall application performance is better.

关 键 词：物联网 跨区域果园 图像增强处理 神经网络 病虫害预警 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]