Rapid Identification Mulberry Harvest Time Based upon Near-Infrared Spectroscopy with the Aid of Mathematical Model
The harvest time of mulberry has a major impact on its medicine quality. In this study, near infrared spectroscopy was used to identify the mulberry harvest time (before the first frost or after the first frost). A total of 85 mulberry samples were randomly divided into calibration set (43) and the prediction set (42). Near infrared diffuse reflectance spectra of mulberry powder were collected, and principal component analysis (PCA), principal components-artificial neural networks (PCs-ANN), principal components-least squares support vector machines (PCs-LS-SVM) were adopted in the identification. The results showed that PCA cannot obtain the goal. The other two methods showed good effects of discrimination; for the calibration set, all can reached prediction accuracy 100％, for the prediction set, all can reached prediction accuracy 97.62％. Among them, PCs-ANN required the minimum number of PCs, just 2; PCs-LS-SVM required 14 PCs. So, it was feasible that a rapid identification of mulberry harvest time based upon near- infrared spectroscopy analysis with the aid of mathematical model.
Near-infrared spectroscopy Mulberry Harvest time Authentic medicine Chemometrics
Hui Yan Shan-shan Mao Shao-qun Li Qiong-ying Wu Bang-xing Han Ming-zhu Jiang Zhong-zheng Gui
School of Biological and Chemical Engineering,Jiangsu University of Science and Technology,Zhenjiang State Key Laboratory of Food Science and Technology, School of Food Science and Technology,Jiangnan
国际会议
The 5th International Congress on Mathematical Biology(第五届国际生物数学大会 ICMB 2011)
南京
英文
1562-1567
2011-06-01（万方平台首次上网日期，不代表论文的发表时间）