Nondestructive determination of GABA in germinated brown rice with near infrared spectroscopy based on wavelet transform denoising

Abstract

The objective of this study was to analyze the content of γ-aminobutyric acid (GABA) in germinated brown rice (GBR) by using near-infrared spectroscopy (NIRS) and the pretreatment method of wavelet de-noising (WD). The prediction accuracy of the NIRS model established by the Daubechies5 wavelet basis function at 3 level denoising treatment is the highest, the correlation coefficient for calibration (rc) is 0.931, the root mean square error of calibration (RMSEC) is 0.4038 mg/100 g, the Bias of calibration is 0.006, the correlation coefficient for prediction (rp) is 0.916, the root mean square error of prediction (RMSEP) is 0.4329 mg/100 g, the Bias of prediction is 0.010, and the ratio of performance to deviation (RPD) is 4.911. Results showed that the predicted and actual values had high correlation. Therefore, these results indicate that the NIRS model treated by WD is feasible to detect GABA content in GBR rapidly and nondestructively.