基于便携式近红外光谱仪和随机森林方法快速鉴别蜂蜜品种
Rapid Identification of Honey Varieties Based on Portable Near-Infrared Spectroscopy and Random Forest Algorithm
乔宁 1饶敏 1黄雪媛 1肖媛 1张岑1
作者信息
摘要
为实现口岸现场快速无损鉴别不同植物源蜂蜜品种,本文以椴树蜜、洋槐蜜、枣花蜜、油菜蜜为研究对象,采用便携式近红外光谱仪与随机森林方法相结合,建立了4 个品种蜂蜜的快速鉴别模型,模型的判别正确率达97.58%.结果表明,便携式近红外光谱仪结合随机森林方法快速鉴别蜂蜜品种,方法准确、快速、无损且高效环保,适用于口岸现场快速分析,可有效满足现场监管需求,为提高通关效率、保障蜂蜜类产品快速通关提供技术参考.
Abstract
To achieve rapid,non-destructive identification of different plant-origin honey varieties at the port of entry and exit,this study focuses on four types of honey:linden honey,acacia honey,jujube blossom honey,and rapeseed honey.A rapid identification model for these honey varieties was established by combining portable near-infrared spectroscopy with the random forest algorithm,with a model discrimination accuracy rate of 97.58%.The results indicate that the method of using a portable near-infrared spectrometer in conjunction with the random forest algorithm for the rapid identification of honey varieties is accurate,fast,non-destructive,and highly efficient and environmentally friendly.It is suitable for rapid analysis at the port of entry and exit,and can provide technical reference for improving customs clearance efficiency,meeting the needs of customs smart control and law enforcement.
关键词
近红外光谱/随机森林/蜂蜜/品种鉴别Key words
near-infrared spectroscopy/random forest/honey/variety identification引用本文复制引用
基金项目
国家重点研发计划(2018YFC0809101)
南昌海关科研项目(2022HK40-06)
出版年
2024
NETL
NSTL
万方数据