近红外光谱技术快速检测小龙虾新鲜度

应用近红外光谱技术实现对小龙虾新鲜度的快速检测。利用化学计量学方法，通过对近红外品质分析仪采集的虾肉绞碎前后光谱（850～1 050 nm）调整不同预处理方法、偏最小二乘法和组合算法，建立一种基于总挥发性盐基氮（total volatile basic nitrogen，TVB-N）含量的小龙虾新鲜度定量预测模型。结果表明：采用标准正态变量变换与一阶导数结合的预处理方法模型预测效果最好，且绞碎后的虾肉光谱比绞碎前建模效果更好；为满足实际应用需要，对绞碎前的虾肉TVB-N含量预测模型进行分析，其交叉验证误差为3.123，交叉验证相关系数为0.947，用此模型对预测集24 个样品进行预测，预测值与实测值的交叉验证相关系数为0.951 4，在TVB-N含量超过20 mg/100 g（不新鲜）的检测准确率为100%。近红外光谱技术可应用于快速检测小龙虾新鲜度，所建模型具有较好的预测能力。

Rapid Detection of Freshness of Crayfish by Near Infrared Spectroscopy

Near infrared spectroscopy (NIRS) combined with chemometrics was used to quickly detect the freshness of crayfish. Near infrared spectra of intact and minced crayfish flesh were recorded in the wavelength range of 850–1 050 nm and preprocessed for the development of a quantitative prediction model for crayfish freshness based on total volatile basic nitrogen (TVB-N) content using partial least square (PLS) and a combinatorial algorithm. The model established using a spectral pretreatment method combining standard normal variate transformation with first derivative had the best prediction performance, and the model based on the spectra of minced crayfish meat had better performance than that developed from the spectra of intact crayfish meat. In order to meet the needs of practical application, the TVB-N content prediction model for minced shrimp meat was analyzed, revealing that the cross-validation error and the cross-validation correlation coefficient were 3.123 and 0.947, respectively. This model was used to predict 24 samples in the prediction set, and it was found that the cross-validation correlation coefficient between the predicted and measured values was 0.951 4, and that the accuracy of prediction was 100% for TVB-N content exceeding 20 mg/100 g (stale samples). In conclusion, NIRS can be used to quickly detect the freshness of crayfish, and the established model has good predictive ability.