عنوان مقاله [English]
Rice is one of the most common and most consumed foods in the world. Quality properties are among the most important factors in determining the cooking and processing characteristics of rice. One of the major problems in the food industry is predicting quality properties. Therefore, in this study, it is possible to predict amylose content (AC), gelatinization temperature (GT), gel consistency (GC), protein content (PC), minimum viscosity (MV), peak viscosity (PV), final viscosity (FV), breakdown viscosity (BDV) and setback viscosity (SBV) of 100 single grain rice samples of Hashemi, Khazar and Dorefak using image processing technology in three exposure treatments including high exposure with LED lamp, high exposure with LED lamp and fluorescent and Anti-light exposure. Calibration models were developed by multivariate linear least squares (PLS) regression. Calibration coefficients of calibration coefficients of variables AC, GT, GC, PC, MV, PV, FV, BDV and SBV in Hashemi, Khazar and Darfak varieties for all treatments were R2cal ≥ 0.89, R2cal ≥ 0.95, and R2cal ≥ 0.92 respectively. prediction coefficients were obtained with R2pre ≥ 0.88, R2pre ≥ 0.94 and R2pre ≥ 0.90, respectively. The results of PLS regression showed that the variables derived from the shape and size characteristics and color variables R,, G, B, L, a and b of image processing were able to predict the rice quality parameters with good accuracy. As a result, using low-cost, non-destructive image processing technology can predict some of the qualitative properties of rice.