摘要
高光谱成像技术因具有快速、客观、范围广、非破坏性等优点,已被广泛应用于食品安全、医学诊断、工业检测等领域。为全面认识高光谱成像技术在粮食品质检测中的应用现状,归纳了高光谱成像技术的基本原理、主要构成及其数据处理与分析,并针对其在粮食品质检测中的应用,重点总结了在粮食检测中水分含量、蛋白质、淀粉及不完善粒的应用研究进展。同时,比较了深度学习与传统的机器学习在粮食品质检测中的应用,探讨高光谱成像技术在样本数据库构建、测量灵敏度、仪器便携化方面所面临的问题与挑战,进而提出改进措施,旨在为粮食的品质评价提供参考。
Hyperspectral imaging(HSI)technology,known for its rapid,objective,extensive,and non-destructive characteristics,has been widely used in fields such as food safety,medical diagnosis,and industrial inspection.In order to comprehensively understand the current application of HSI technology in grain quality detection,this paper summarizes the basic principle,main composition,data processing and analysis of HSI,Focusing on its application in grain quality detection,this paper highlighted the research progress on the application of HSI technology for detecting moisture content,protein,starch and imperfect grains in grain detection.Additionally,this paper compared the application of deep learning and traditional machine learning in grain quality detection,discussed the problems and challenges faced by HSI technology in sample database construction,measurement sensitivity,and instrument portability,and then proposed improvement measures,aiming to provide references for grain quality evaluation.
作者
李彭
李艳艳
何学明
刘强
邢常瑞
方勇
袁建
Li Peng;Li Yanyan;He Xueming;Liu Qiang;Xing Changrui;Fang Yong;Yuan Jian(School of Food Science and Engineering,Nanjing University of Finance and Economics/Collaborative Innovation Center of Modern Grain Circulation and Security in Jiangsu Province,210023)
出处
《粮食储藏》
2024年第4期1-12,共12页
Grain Storage
基金
国家重点研发计划课题(2021YFD2100601)
江苏省研究生科研与实践创新计划项目(KYCX24_1937)。
关键词
高光谱成像技术
粮食品质
水分含量
蛋白质
淀粉
不完善粒
深度学习
hyperspectral imaging technology
grain quality
moisture content
protein
starch
imperfect grain
deep learning