贮藏期内富士和粉红女士苹果品质的FT-NIR无损检测

Non-Destructive Quality Evaluation of Fuji and Pink Lady Apples by FT-NIR during Storage Period

为探索傅里叶近红外光谱快速无损检测贮藏期苹果品质的方法,在苹果贮藏过程中,每隔30d采集富士和粉红女士(各40个)2个苹果品种共计400个样本的近红外图谱(12000～4000cm-1),用OPUS-QUANT软件预处理光谱,用偏最小二乘法建立通用于2个品种的可滴定酸(TA)、pH值和可溶性固形物(SSC)的数学模型。结果表明:富士和粉红女士的光谱经矢量归一化预处理后,在波段7502～4247cm-1内所建立的可滴定酸模型稳定性较好,该模型校正时的相关系数(R2)和评估均方误分别为0.9231和0.0263%,预测时的相关系数R2和内部交叉验证均方根差分别为0.9071和0.0266%;在波段11995～4247cm-1内,光谱经一阶导数预处理后所建立的pH值预测模型稳定性较好,该模型校正时的R2和评估均方误分别为0.9263和0.0700,预测时的R2和内部交叉验证均方根差分别为0.9113和0.0772;近红外光谱经最大-最小归一化预处理后,在波段6102～5446cm-1所建立的SSC模型效果较好,该模型校正时的R2和评估均方误分别为0.9212和0.3570%,预测时的R2和内部交叉验证均方根差分别为0.9130和0.370%。在富士和粉红女士贮藏期品质检测过程中,建立的通用于这2个品种的TA、pH值和SSC检测的数学模型,稳定性较好,能满足品质快速无损检测的要求。

In order to explore the applicability of FT-NIR technique for rapid and non-destructive evaluation of apple quality in terms of titratable acidity（TA）,pH and solid soluble content（SSC）,NIR spectra in the wavelength range of 12000 to 8000 cm-1 were acquired from 40 samples including Fuji and Pink Lady apples during the storage period of 30 days.Universal mathematical models of TA,pH and SSC for both apple varieties were established using partial least square（PLS） regression.The results showed that a stable model of TA was developed in the wavelength range of 7502 to 4247 cm-1.The coefficient of correlation（R2） of calibration and root mean square error of estimation（RMSEE） were 0.9231 and 0.0263%,respectively.Meanwhile,the coefficient of correlation of prediction（R2） and the root mean square error of cross-validation（RMSECV） were 0.9071 and 0.0266%,respectively.In addition,a stable model of pH was achieved by PLS＋FD model based on NIR spectra in the wavelength range of 11995 to 4247 cm-1.The R2 of calibration and RMSEE were 0.9263 and 0.0700,respectively,and the R2 of prediction and RMSECV were 0.9113 and 0.0772,respectively.A good model of SSC was obtained by min-max normalization pretreatment in the wavelength range of 6102 to 5446 cm-1.The R2 of calibration and RMSEE were 0.9212 and 0.3570%,respectively,and the R2 of prediction and RMSECV were 0.9130 and 0.370%,respectively.The models of TA,pH and SSC were stable in evaluating the quality of Fuji and Pink Lady apples and could meet the requirements for rapid and non-destructive evaluation of fruit quality.