基于高光谱的牛奶脂肪质量浓度预测模型建立与评价

Prediction and validation of fat content in milk based on hyperspectrum

以牛奶中脂肪质量浓度的检测为研究对象,应用图像处理技术分析高光谱数据,选取样品分析区域,提取分析区域的平均光谱,建立了PLS、N-PLS预测模型对牛奶中脂肪质量浓度进行分析。结果表明,PLS预测模型中校正集与预测集的相关系数分别为0.9851和0.9913,而N-PLS预测模型中校正集与预测集的相关系数分别为0.9999和0.9976。两个模型对比发现,N-PLS模型预测精度高于PLS模型,尤其是牛奶中脂肪质量浓度较少的情况。

In this paper, the detection of fat content in milk was studied. The image processing technology was used to select the sample area, and the average spectrum was extracted. The PLS and N-PLS prediction model were established to analyze the fat content in milk. The re- suits showed that the correlation coefficients between the calibration set and the prediction set in the PLS prediction model were 0.9851 and 0.9913 respectively, and the correlation coefficients between the calibration set and the prediction set in the N-PLS prediction model were 0.9999 and 0.9976 respectively. The accuracy of N-PLS model was higher than that of PLS model, especially in the case of low fat in milk.