Rapid identification of the variety and geographical origin of Wuyou No.4 rice by fourier transform near-infrared spectroscopy coupled with chemometrics

Rapid and non-destructive discriminant analysis of paddy rice spectra has great significance for large-scale agriculture, including rice seed identification, grain acquisition, factory processing, etc. Here, we describe a rapid identification method for the varietal and geographical origin of commercial rice grain using Fourier transform near-infrared (FT-NIR) diffuse-reflectance spectroscopy coupled with principal component analysis (PCA) and deep learning (DL) methods. We applied the methods to identify a high-quality rice variety WuYou No. 4 from other rice varieties having similar rice grain shapes. The total recognition-accuracy rates for sample calibration and testing by PCA method was found to reach 91.04% and 87.10%, respectively, whereas both reached 100% by applying the DL method. When considering identification of the geographical origin from the Wuchang geographical area, DL model results showed excellent discriminant power with total classification and testing rates of 93.31% and 96.67%, respectively. The category rates for Wuyou No.4 samples from Wuchang area and other area are 86.82% and 98.61%, respectively. Our results demonstrate the high potential of FT-NIR spectroscopy as a fast, nondestructive and environmentally safe method for the rapid and reliable identification of the variety and geographical origin of commercially important high quality Wuyou No.4 rice.