LS-SVM for Quantification Soluble Solids Content of Intact Apples by Transmittance Spectroscopy
The feasibility of quantification soluble solids content (SSC) of intact apples was investigated by visible and near infrared (Vis-NIR) transmittance spectroscopy combined with LS-SVM method. The spectra were pretreated by Savitzky-Golay smoothing, first derivative and second derivative. The regression models were developed by least squares support vector machines (LS-SVM) and partial least squares (PLS). The accuracy of the LS-SVM models and PLS models were compared. The performance of models was better by LSSVM method with higher correlation coefficients (r) of 0.98, and lower standard error of prediction (SEP) of 0.33°Brix. The results showed that Vis-NIR transmittance spectroscopy using LS-SVM technique could improve the precision of quantification SSC of intact apples nondestructively.
Visible and near-infrared spectroscopy least squares support vector machines apple soluble solids content
Yande Liu Yanrui Zhou
East China Jiao tong University Institute of Optics-Mechanics-Electronics Technology and Application (OMETA) Nanchang, China
国际会议
哈尔滨
英文
804-808
2012-06-16（万方平台首次上网日期，不代表论文的发表时间）