99B-26 |
|
M. MONSOOR, Department of Food Science, University of Arkansas, 2650 N Young Avenue, Fayetteville, AR 72704 and A. Proctor, Food Science, Univ. of Arkansas, 2650 Young Ave., Fayetteville, 72704. Residual rice bran is the major source of milled rice surface lipid that affects the stability, quality, and value of milled rice. Common physical methods to determine residual bran are based on the amount of light reflected from, or transmitted through, milled rice that depend on the appearance of milled rice. However, there are no reports of surface lipid content measurement of milled rice using a simple Hunter Color Instrument. The objectives of this research were 1) to develop calibration models by relating surface lipid content with the whiteness, degree of milling (DOM), Hunter ‘L’ values, and Hunter ‘+b’ values of milled rice and 2) to validate the calibration models using an independent sample set. Calibration models were developed with one independent (Y) and one response (X) variable where X is the total surface lipid contents, Y is one of the independent variables (DOM values, whiteness values, Hunter ‘L’ values, or Hunter ‘b’ values). Surface lipid contents of 15 rice samples of various residual bran contents were predicted from the linear fit equation of the calibration model and correlated with the calculated values to determine the accuracy of the models. The predicted milled rice surface lipids using whiteness, DOM, Hunter ‘L’ values, and Hunter ‘+b’ color values correlated well with the measured values obtained by chemical analysis. The standard error of prediction and R2 values indicated that Hunter ‘+b’ value was the best predictor for surface lipid contents. The results of this research suggested that the surface lipid contents of the milled rice could be calculated by Hunter ‘+b’ values. Therefore, Hunter Color Instrument has the potential to replace the conventional milling meter for online evaluation of milled rice surface lipid.
Session 99B, Food Engineering: Process analysis, control and modeling
|