Development of Simplified Models for Nondestructive Testing of Rice with Husk Starch Content Using Hyperspectral Imaging Technology
The study aimed to establish predictive models of starch content in rice (with husk) using a hyperspectral imaging system (HSI) in a collection of 87 different rice varieties in China in the wavelength range of 938-2215 nm and the first established multivariate calibration models over the full wavelength range by using partial least-squares regression (PLSR), principal component regression (PCR) and support vector machines (SVM). In the process of predictive model optimization, the optimal wavelengths were selected by using regression coefficients as discriminating factors to establish PLSR models and according to the regression coefficient and root mean square error of models, the optimal wavelengths were filtered from 7 to 5. The model based on the seven optimal wavelengths was determined as the optimal optimization model with a R2P of 0.8029 and a RMSEP of 1.79. The mapping of starch content was achieved by transferring a quantitative model to each pixel. According to the visualization image, the distribution of rice starch could be understood, thus realizing the possibility of on-line detection of starch content by using hyperspectral imaging technology.