|国家科技期刊平台
首页|期刊导航|农机化研究|基于深度学习的植保无人机识别系统应用研究

基于深度学习的植保无人机识别系统应用研究OA

Application of Identification System for UAV Plant Protection Based on Deep Learning

中文摘要英文摘要

针对人工识别病虫害种类并定位的方法精度低、速度慢且人工成本巨大的问题,基于深度学习对植保无人机的识别系统应用进行研究.该识别系统主要包括 3 层,分别是感知层、传输层和应用层.为使植保无人机能够准确地识别病虫害种类并定位,采用深度学习算法进行人工智能识别,并对卷积神经网络和识别的评价指标进行了设计.为验证该识别系统的性能,进行了训练和测试试验及系统在其他数据集的测试试验.试验结果表明:系统的病害识别率较高,且具有良好的可移植性.

Aiming at the problem of low accuracy,slow speed and huge labor cost of manual identification and positio-ning of pest species,the application of identification system of plant protection UAV based on deep learning was studied.The recognition system was mainly constituted of three layers:perception layer,transmission layer and application layer.Deep learning algorithm was used for artificial intelligence recognition to enable plant protection UAV to accurately identi-fy and locate pest specied.The convolutional neural network and recognition evaluation indicators ware designed.To verify the performance of the recognition system,training and test,test of model in other data sets were carried out.The test re-sult show that the system has high virus recognition rate and good portability.

张新英;林婷婷;付川南

郑州经贸学院,郑州 451191

农业工程

植保无人机识别系统深度学习卷积神经网络识别率

plant protection UAVidentification systemdeep learningconvolutional neural networkrecognition rate

《农机化研究》 2024 (005)

182-186 / 5

河南省骨干教师培养计划项目(2018GGJS213);河南省高等学校重点科研项目(22B520043);郑州经贸学院骨干教师项目(ggjs1902);河南省精品在线开放课程建设项目(2021);河南省本科一流课程建设项目(2022);2021郑州经贸学院一流课程建设项目(ylkc08)

评论