论文部分内容阅读
应用近红外反射光谱技术(NIRS)对滇南小耳猪热鲜均质肉糜和绝干粉的水分、粗脂肪、粗蛋白含量进行建模研究,并筛选出最优的光谱预处理方法。采集11 000~4 300 cm-1范围内43份猪肉样品光谱数据,在多元散射校正(MSC)、二阶导数(Second derivative)、变量标准化校正(SNV)不同组合方式的光谱预处理基础上,采用偏最小二乘法(PLS),建立滇南小耳猪猪肉的水分、粗脂肪、粗蛋白质3个化学组分的近红外预测模型,筛选最佳的光谱预处理方法和主成分数。水分预测较好的是匀质肉糜原始光谱预测,R2为0.981,RMSEC为0.177,RMSEP为0.810,最佳主成分数为7;粗脂肪和粗蛋白预测效果较好的均是绝干粉的原始光谱,R2分别为0.986、0.976,RMSEC分别为0.567、0.765,RMSEP分别为2.325、2.697,最佳主成分数均为7。因此,近红外光谱分析方法能够很好地检测滇南小耳猪猪肉中的水分、粗脂肪和粗蛋白。
Near infrared reflectance spectroscopy (NIRS) was used to study the moisture content, crude fat and crude protein content in the dried minced fresh minced pork meat and dry powder, and the optimal spectral pretreatment method was screened out. Spectral data of 43 pork samples in the range of 11 000 ~ 4 300 cm-1 were collected. Based on the spectral pretreatment of different combinations of multiple scattering calibration (MSC), second derivative and variable standardization correction (SNV) The partial least squares (PLS) method was used to establish the near infrared prediction model of moisture, crude fat and crude protein in the pork of South Yunnan small ear pigs. The optimal spectral pretreatment methods and principal component numbers were screened. Prediction of moisture content is better than the original spectrum of homogenized meat emulsion, R2 is 0.981, RMSEC is 0.177, RMSEP is 0.810, the best principal component is 7. The predicted results of crude fat and crude protein are better than the original spectra of absolute dry powder , R2 respectively 0.986,0.976, RMSEC respectively 0.567,0.765, RMSEP respectively 2.325,2.697, the best principal components are all 7. Therefore, near-infrared spectroscopy can detect the moisture, crude fat and crude protein in the pork of Diannan small-ear pigs well.