文档介绍：: .
中国粮油学报特征融合的玉米籽粒霉变等级分类
模型。经分析比较，融合光谱和图像特征并结合 ELM 算法建立的分类模型用于玉米籽粒霉
变等级识别效果最优，训练集和测试集分类准确率(Acc)分别为 %和 %，并将玉米
籽粒霉变等级进行可视化表达。研究结果有望为玉米霉变籽粒高光谱现场快速检测设备研发
提供技术支持。
关键词：玉米霉变籽粒、高光谱成像技术、随机蛙跳、极限学习机
中图分类号： 文献标识码：A 网络首发时间： 2022-05-17 10:19:58
网络首发地址： .
Study on Hyperspectral Image Detection
基金项目: 中央级公益性科研院所基本科研业务费专项资金（ZX1937）
收稿日期：2021-11-09
第一作者：杨东，男，1987 年出生，博士，粮食储藏技术与品质检测
通信作者：石天玉，男，1980 年出生，副研究员，储粮生态理论与工艺
1中国粮油学报
Method of Maize Grain Mildew Grade
Yang Dong1 Wang Shuhui2 Wu Jianhua2 Jiang Junyi1 Song Kai2 Shi Tianyu1
（Academy of National Food and Strategic Reserves Administration, National Engineering
Research Center of Grain Storage and Logistics1, Beijing 100037）
（Shenyang Ligong University2, Shenyang 110159）
Abstract: In order to quickly and undamaged detect the different mildew conditions of stored corn
grains and improve the quality inspection efficiency of corn collection and storage, the paper
attempts to build classification models of corn grain mildew by using hyperspectral imaging
technology combined with machine learning algorithm. The hyperspectral images of corn grains
in the range of 400-1000nm were collected. Based on the measured number of fungal spores, the
grain mildew state was divided into four grades: healthy, mild, moderate and severe mildew.
Seven spectral characteristic variables were optimized by random frog (RF) algorithm. For the
characteristic band images, 21 texture characteristic variables we