采用近红外光谱法无损测量樱桃品质的研究

Fruit Quality Indexes Nondestructive Prediction of Cherry Using Near-infrared Spectroscopy

以樱桃常见品种‘雷尼尔’、‘布鲁克斯’、‘雷吉纳’、‘美早’为试材,采用近红外光谱(NIR)对可溶性固形物含量、干物质含量、果皮颜色和果实硬度4个质量指标进行无损测量,开发预测性NIR回归模型,并使用该模型测量‘雷尼尔’樱桃从果实开始着色到完全成熟阶段4个质量指标变化。结果表明,建立的可溶性固形物含量、干物质含量、果皮颜色和果实硬度模型的线性回归决定系数(R^(2))值分别为0.87、0.97、0.77、0.76。模型线性系数(R2)越接近1拟合效果越好,模型的拟合程度越高。R^(2)值均大于0.7,说明模型能够达到性能预期。可溶性固形物和干物质模型进行定量测定的准确性高于果皮颜色和果实硬度模型。樱桃果实成熟阶段,可溶性固形物与干物质逐渐增加,果皮颜色逐渐加深,硬度逐渐降低。本研究建立的模型可用于其他樱桃品种品质预测,同时可为进一步探索樱桃果实成熟阶段内部物质变化提供参考依据。

Soluble solids content,dry matter,peel color and firmness of cherry‘Rainier’,‘Brooks’,‘Regina’,‘Tieton’were measured with no damage by near-infrared spectroscopy(NIR)in the experiment.Predictive NIR regression model was developed to measure four quality indicators using the‘Rainier’cherry fruit from beginning of coloration to fully maturity.The result showed that the coefficients of determination(R^(2))of soluble solids content,dry matter,peel color and firmness were 0.87,0.97,0.77 and 0.76,respectively.The closer the model linearity(R^(2))of the model gets to 1,the better the fitting effect and the higher the fitting degree of the model.The R^(2) values were all greater than 0.7,indicating that the model could achieve the performance expectations.The accuracy of the soluble solids model and dry matter model was higher than that of models of peel color and firmness.The soluble solid and dry matter gradually increased,skin color gradually deepened and hardness gradually decreased during the ripening of cherry fruit.The model established in this study can be applied in predicting four quality indexes for other cultivars,providing the intuitive reference for exploring changes on endogenous substances in the fruit ripening stage.