论文部分内容阅读
利用黑龙江省、吉林省、辽宁省和江苏省的88份常规粳稻品种,建立了适合中国粳稻食味评价的可见光-近红外光谱预测模型。感官食味试验采用小样品感官试验综合评分法,其相关性分析和逐步回归结果表明,感官综合评分与气味、色泽、外观完整性、味道、口感均呈极显著正相关。通过粳稻感官食味评分值与可见光-近红外光谱值的回归分析,得到7个自变量(4个波长的吸光度、碎米率、裂纹米率和香气系数)的预测回归方程,其相关系数R均达到0.699~0.799的极显著相关,经验证该预测回归方程能够较好的预测粳稻的米饭食味品质。
Using 88 conventional japonica rice varieties in Heilongjiang, Jilin, Liaoning and Jiangsu Provinces, a prediction model of visible-near-infrared spectra suitable for the evaluation of Chinese japonica rice taste was established. Sensory taste test using small sample sensory test comprehensive score method, the correlation analysis and stepwise regression results show that the sensory comprehensive score and the smell, color, appearance integrity, taste, taste were extremely significant positive correlation. Through regression analysis of sensory taste score and visible-near-infrared spectral value of japonica rice, predictive regression equations of seven independent variables (absorbance, broken rice rate, cracked rice rate and aroma coefficient at four wavelengths) were obtained. The correlation coefficients R Reaching a significant correlation of 0.699 ~ 0.799. It is verified that the prediction regression equation can better predict the rice taste quality of japonica rice.