基于语义分割的农作物病害识别综述研究  

Survey of Crop Disease Recognition Based on Semantic Segmentation

作　　者：赵小虎[1,2] 李晓 ZHAO Xiaohu;LI Xiao(National and Local Joint Engineering Laboratory of Internet Application Technology on Mine,(China University of Mining and Technology),Xuzhou 221008;School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221008)

机构地区：[1]矿山互联网应用技术国家地方联合工程实验室(中国矿业大学),徐州221008 [2]中国矿业大学信息与控制工程学院,徐州221008

出　　处：《现代计算机》2021年第18期122-126,共5页Modern Computer

基　　金：徐州市重点研发科技项目(No.KC19112)。

摘　　要：计算机视觉是在感知层上最为重要的核心技术之一,被应用于众多领域。在农业领域,主要应用于植物生长监测、农作物病害的监测与防治等。结合深度学习发展的大背景,将图像语义分割分为基于编-解码架构和基于融合架构,并对优秀学者在农作物病害识别方面的研究进行详细介绍,总结农作物病害图像分割应用中所存在的问题。Computer vision is one of the most important core technologies at the perceptual level and is used in many fields.In the agricultural field,it is mainly used in plant growth monitoring,crop disease monitoring and prevention,etc.Combining the background of deep learning devel⁃opment,image semantic segmentation is divided into encoding-decoding architecture and fusion architecture,and the research of outstand⁃ing scholars in crop disease recognition is introduced in detail,and the problems existing in the application of crop disease image segmenta⁃tion are summarized.

关 键 词：图像语义分割 深度学习 农作物病害识别 监督学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S432[自动化与计算机技术—计算机科学与技术]