近红外光谱技术快速测定鹅肉嫩度

目的：应用近红外光谱技术快速检测鹅肉的嫩度值。方法：采集完整鹅肉的近红外光谱（950～ 1 650 nm），光谱经多种校正预处理后，再分别采用主成分回归和偏最小二乘法建立鹅肉嫩度的定量预测数学模 型。结果：采用5点移动窗口平滑处理结合偏最小二乘法所建立模型的预测效果最好，嫩度定量校正数学模型的模 型决定系数为0.908 0，内部交互验证均方根误差为113.618 6。用此模型对预测集20 个样品进行预测，预测值与实 测值的相关系数达到0.971 1，预测值平均偏差为21.673 g，预测值和实测值之间没有显著性差异（P＞0.05）。结 论：近红外光谱作为一种无损快速的检测方法，可用于评价鹅肉的嫩度。

Rapid Determination of Goose Tenderness Using Near Infrared Spectroscopy

Objective: To propose a rapid method for the identification of goose tenderness by near infrared spectroscopy
(NIR) technology. Methods: NIR spectra (950–1 650 nm) of goose meat were collected. After multiple correction and
pretreatment of the spectra, mathematical models for the quantitative prediction of goose tenderness were established by
principal component regression (PCR) and partial least squares regression (PLSR). Results: The PLSR model based on
five-point moving window smoothing was the best predictive model with a determination coefficient (R2) of 0.908 0 and
a root mean square error of cross validation (RMSECV) of 113.618 6. No significant difference (P ＞ 0.05) was found
between the predicted and measured values for 20 samples in the prediction set, with a correlation coefficient of 0.971 1,
and the predicted values showed an average bias of 21.673 g. Conclusion: NIR can be used in the evaluation of goose meat
tenderness as a fast nondestructive detection method.