基于近红外漫反射光谱检测鸡蛋品质

对利用傅里叶变换近红外漫反射光谱快速检测鸡蛋品质和鸡蛋存储时间的可行性进行了探讨。运用偏最小二乘法(PLS)构建了用于鸡蛋哈夫单位、蛋白pH值、蛋形指标等鸡蛋品质和存储时间的光谱模型，采用交互验证决定系数R2cv，交互验证均方根误差(RMSECV)对模型精度进行了评价。结果表明，模型经优化后哈夫单位、蛋白pH值、蛋形指标和存储时间的光谱模型的R2cv分别为0.86、0.84、0.26和0.92，RMSECV分别为7.52、0.17、0.05和1.37。这些结果表明除蛋形指标外，模型具有较好的准确性和预测能力，能够满足鸡蛋品质新鲜度快速无损检测的要求。随着储存时间的变化，除了蛋形指标无变化外，其它各项指标与储存时间都存在较高的相关性，符合新鲜鸡蛋变化趋势。

Detection of egg quality by near infrared diffuse reflectance spectroscopy

The feasibility of using the Fourier transform near-infrared diffuse reflectance spectrometry to determine egg quality and storage time rapidly was investigated. The spectral models were established based on Partial Least Square(PLS) for egg quality indexes(Haugh unit, protein pH and egg shape) and storage time and the coefficient of determination(R2cv) and the Root Mean Square Error of Cross Validation(RMSECV) were used to evaluate the model accuracy. The results show that the optimal predictive models for Haugh unit, protein pH, egg shape and storage time are obtained by effective information selection with the R2cv of 0.86、0.84、0.26 and 0.92, respectively and the RMSECV of 7.52, 0.17, 0.05 and 1.37, respectively. It concludes that the models have better measuring accuracy and prediction capabilities except for the egg shape, which can meet the requirements of fast non-destructive detection for the fresh eggs. Moreover, All parameters have a higher correlation with the change of storage time except for the egg shape, which meets the change tendency of the storage fresh eggs.