近红外光谱无损检测石榴中花色苷含量研究

以全国不同地区的97个石榴为样本,研究近红外光谱无损检测石榴中花色苷的含量,探讨了不同数据处理和回归方法对建模效果的影响。结果表明,对原始光谱进行一阶微分、标准多元离散校正法处理后,采用偏最小二乘法建立的石榴花色苷含量预测模型,预测偏差为0.148,预测标准差(SEP)为1.47,相关系数为0.829,模型预测良好,说明近红外光谱无损检测石榴的品质是可行的。

Non-destructive Determination of Anthocyanins in Pomegranate by Near Infrared Spectroscopy(NIRS)

With 97 pomegranates from different regions as the research samples,the non-destructive determination of anthocyanins content in pomegranate by near infrared spectroscopy is studied,and the effects of different data processing and regression methods on modeling are discussed.The prediction model is established by first derivative treatment with standard multi scatter correction(MSC)and partial least square(PLS).The results show that model provides the better predication of anthocyanins in pomegranate with the bias of 0.148,the standard error of prediction(SEP)of 1.47 and the correlation coefficient of 0.829.Research results suggest that NIRS technique is feasible for non-destructive determination of pomegranate quality.