近红外光谱法快速检测带鱼肉中的水分和蛋白质含量

应用傅立叶变换近红外光谱分析技术对带鱼肉中水分和蛋白质含量进行了研究。分别建立原始光谱、间隔2点一阶导数(dblg2)、3点平滑(sa3)、标准归一化(SNV)和多元散射校正(MSC)的偏最小二乘回归(PLS)模型,比较定标相关系数(Rc)、预测相关系数(Rv)、定标标准差(SEC)和预测标准差(SEP),建立了MSC预测模型,水分和蛋白质近红外检测模型的相关系数均在0.9以上。SEC分别为0.74和0.68,SEP分别为0.81和0.73。将确定的模型进行了外部验证,水分和蛋白质NIR预测值和化学分析值的配对t检验差异均不显著。说明近红外光谱法应用于带鱼肉中水分和蛋白质含量的快速检测是可行的。

Rapid determination of moisture and protein contents in hairtail mince by near infrared reflectance spectroscopy(NIRS)

Moisture and protein content of hairtail mince were determined by fourier transform near infrared reflectance spectroscopy(FT-NIRS).The partial least squares(PLS) prediction models were developed based on the original,dblg2,sa3,SNV and multiplicative scatter correction(MSC) spectra,respectively.By comparing the Rc,Rv,SEC and SEP,the MSC model was established.The correlation coefficients of the moisture and protein calibration models obtained were above 0.9,the SEC was 0.74 and 0.68,respectively,the SEP was 0.81 and 0.73,respectively.The models were further validated,the result showed that there was no significant difference between NIR prediction value and chemical analysis value.So the NIR spectroscopy technology was a rapid and effective technique which was suitable to determine moisture and protein content in hairtail mince.