基于自主飞行器的柑橘病害监护系统设计与实践  

Design and Practice of Citrus Disease Monitoring System Based on Autonomous Aircraft

作　　者：余焕杰 江华晋 彭东海 YU Huanjie;JIANG Huajin;PENG Donghai

机构地区：[1]韶关学院信息工程学院,广东韶关512005

出　　处：《智慧农业导刊》2024年第11期1-5,共5页JOURNAL OF SMART AGRICULTURE

基　　金：广东省省级大学生创新创业训练计划项目(S202310576020X);韶关市科技局项目(220607154531465)。

摘　　要：针对人工手段下的柑橘病害监护过程中存在的效率低下与成本巨大的问题,基于自主飞行器对柑橘病害监护系统进行设计与实践。该监护系统主要由感知层、应用层、分析层3个部分组成。为了使飞行器可以准确地识别柑橘病害,利用DenseNet121与EfficientNetB7深度学习模型框架训练出专用于识别柑橘病害的柑橘病害识别模型,并将其部署在位于自主飞行器的香橙派开发板上,从而为自主飞行器提供识别柑橘病害的能力。为了验证该病害监护系统的性能,进行测试试验。试验结果表明,该病害监护系统具有良好的发展潜能,模型拥有良好的检测精度与可移植性。Aiming at the problems of low efficiency and huge cost in the process of citrus disease monitoring by manual means,a citrus disease monitoring system is designed and practiced based on an autonomous aircraft.The monitoring system is mainly composed of three parts:perception layer,application layer and analysis layer.In order to enable the aircraft to accurately recognize citrus diseases,a citrus disease recognition model dedicated to recognizing citrus diseases was trained using the DenseNet121 and EfficientNetB7 deep learning model framework,and deployed on the Aroma Orange Pi development board located in the autonomous aircraft,thus providing the autonomous aircraft with the ability to recognize citrus diseases.In order to verify the performance of the disease monitoring system,tests were carried out.The experimental results show that the disease monitoring system has good development potential,and the model has good detection accuracy and portability.

关 键 词：智慧农业 深度学习 图像识别 飞行器设计 病害监护系统 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]