Premium
Discrimination of sesame oil adulterated with corn oil using information fusion of synchronous and asynchronous two‐dimensional near‐mid infrared spectroscopy
Author(s) -
Yang Renjie,
Dong Guimei,
Sun Xueshan,
Yang Yanrong,
Liu Haixue,
Du Yanhong,
Jin Hao,
Zhang Weiyu
Publication year - 2017
Publication title -
european journal of lipid science and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.614
H-Index - 94
eISSN - 1438-9312
pISSN - 1438-7697
DOI - 10.1002/ejlt.201600459
Subject(s) - linear discriminant analysis , pattern recognition (psychology) , fusion , artificial intelligence , support vector machine , mathematics , near infrared spectroscopy , principal component analysis , spectroscopy , asynchronous communication , computer science , physics , optics , computer network , philosophy , linguistics , quantum mechanics
A new and accurate approach for classifying sesame oil adulterated with corn oil using the information fusion of synchronous and asynchronous 2D NIR‐MIR correlation spectroscopy combined with least squares support vector machines (LS‐SVM). N‐way partial least square discriminant analysis (NPLS‐DA) was adopted to extract the principal components of synchronous and asynchronous 2D NIR‐MIR correlation matrices. And the score matrices of synchronous and asynchronous 2D NIR‐MIR correlation spectroscopy were emerged. Both separate and emerged score matrices were input LS‐SVM to construct discriminate models for adulterated sesame oil. The ratios of correct classification were 98.1 and 100% for calibration set and prediction set, respectively, based on information fusion combined with LS‐SVM. Comparison results showed that the LS‐SVM model could provide higher discriminant accuracy for adulterated sesame soil based on the information fusion of synchronous and asynchronous 2D NIR‐MIR correlation spectroscopy than both separate spectra information. Practical applications: The proposed method not only contains NIR and MIR spectral information, but also contains the similar and dissimilar variation spectral information with the perturbation. Thus, the method can provide better discriminate results of adulterated sesame oil in comparison with conventional spectral methods; can also be applied to other food safety detection areas. The information fusion of synchronous and asynchronous 2D NIR‐MIR correlation spectroscopy is proposed for classifying adulterated sesame oil.