摘要
利用近红外光谱分析仪采集2012年度江苏、安徽、河南等省份25份农户田间小麦品种籽粒样品的近红外光谱信息,对获取的近红外光谱数据分别进行均值标准化、一阶求导、二阶求导和多元散射校正处理,利用全波段(950~1 650nm)和特征波长处(985、1 130、1 160、1 190、1 235、1 320、1 385、1 410 nm)的近红外光谱数据,采用离差平方和法(Ward法)聚类分析和主成分分析等化学计量学方法,构建赤霉病小麦籽粒和未病小麦籽粒的SIMCA识别模型.模型诊断和验证结果显示,构建的SIMCA识别模型对赤霉病小麦籽粒和未感病小麦籽粒的正确识别率均为100%,识别效果良好.
A total of 25 wheat kernel samples collected in 2012 from Jiangsu,Anhui and Henan Province of China were analyzed by near infrared reflectance spectroscopy(NIRS).After normalization,first derivative,second derivative and multiplicative scattering correction treatment were conducted for wheat kernel samples spectral data,using the total spectra area(950?1650 nm)and NIR data in the characteristic wavelength area(985,1130,1160,1190,1235,1320,1385,1410 nm),systematic cluster analysis and principal component analysis of sum of squares of deviations were applied to build the SIMCA identification model to classify the wheat kernel samples infected or not by Fusarium Graminearum Schw.Model diagnosis and validation results showed that the correct identification rate of SIMCA identification model to classify the wheat kernel samples infected or not by Fusarium Graminearum Schw was 100%,and identification effect was excellent.
作者
关二旗
崔贵金
卞科
郑祝红
Guan Erqi;Cui Guijin;Bian Ke;Zheng Zhuhong(College of Food Science and Technologe,Henan University of Technology,Zhengzhou 450001;Henan Food Crop Collaborative Innovation Center,Zhengzhou 450001;Yihai Kerry(Quanzhou)Oil,Grain&Foodstuffs Industries Co.,Ltd,Quanzhou 362804)
出处
《中国粮油学报》
EI
CAS
CSCD
北大核心
2019年第S01期71-76,共6页
Journal of the Chinese Cereals and Oils Association
基金
国家自然科学基金(31401523)
国家粮食公益性行业科研专项(201313005-04)
国家现代农业(小麦)产业技术体系专项(CARS-03).
