基于嵌入式田间杂草识别及定位方法探究  

Exploration of Weed Identification and Localization Method Based on Embedded Field

作　　者：王云德[1] 刘孜文[1] 徐生龙[2] WANG Yunde;LIU Ziwen;XU Shenglong(Gansu Polytechnic College of Animal Husbandry&Engineering,Wuwei 733006,China;Wuwei Vocational College,Wuwei 733000,China)

机构地区：[1]甘肃畜牧工程职业技术学院,甘肃武威733006 [2]武威职业学院,甘肃武威733000

出　　处：《农机使用与维修》2023年第6期93-95,共3页Agricultural Machinery Using & Maintenance

摘　　要：田间除草是农业生产中的重要环节,其除草效率直接影响田间作物产量。为了提高田间除草效率与智能化水平,基于深度学习的方法,实现田间杂草的精准识别与定位,并构建嵌入式计算平台,融合深度学习训练所得的模型,实现田间杂草的精准判断、定位与去除,可以提高田间除草作业的智能化水平,降低劳动强度与环境污染。Field weeding is an important link in agricultural production,and its weeding efficiency directly affects the yield of crops in the field.In order to improve the efficiency and intelligence level of field weed control,based on deep learning methods,precise identification and positioning of field weeds are achieved,and an embedded computing platform is constructed.The model trained through deep learning is integrated to achieve precise judgment,positioning,and removal of field weeds.This can improve the intelligence level of field weed control operations,reduce labor intensity and environmental pollution.

关 键 词：深度学习 田间除草 定位 嵌入式 智能 

分 类 号：S451.1[农业科学—植物保护]