基于近红外光谱技术的雷竹笋品质指标快速检测方法研究

Rapid Determination Method for Quality Parameters of Bamboo Shoot by Near Infrared Spectroscopy

采用傅里叶变化近红外光谱技术建立雷竹笋的硬度、色差和水分含量快速检测方法。选取4 000~8 500cm-1为分析区域,描述谱峰的归属。以常规分析测定值作建模数据,采用偏最小二乘（PLS）回归法建立雷笋的品质参数定量分析模型,考察近红外光谱预处理方法对模型的影响。分别用验证集和校正集样本分析模型预测的准确性。结果显示预测集均方差（RMSEP）分别为150、1.8、0.0012,校正集均方差（RMSEC）分别为413、2.3、0.0042,相关系数分别为0.985、0.940、0.988。对传统方法和近红外光谱法进行t检验,硬度、色差和水分的t值分别为1.93、0.68和1.329,均小于临界值t0.05=2.131,说明两者之间没有显著性差异。本研究结果表明,采用近红外光谱法可同时测定雷笋的硬度、色差和水分含量等品质参数。与传统的分析方法相比,该方法具有快速、无损、简单等特点。

A rapid determination method of the quality parameters such as texture, chromatic aberration and mois ture content in bamboo shoot by FT-near infrared spectroscopy was proposed in this study. The spectrum interval of 4 000-8 500 cm-1 was selected as the analysis spectrum. Partial Least-squares Regression （PLSR） models of quality pa- rameters were established based on the data by traditional method, and the effect of spectrum pretreatment method on the prediction ability of the models was investigated, and the precise of the models was also evaluated with the calibra- tion and validation sets. The results showed that the root mean square error of perdition set （RMSEP） for the models of texture, chromatic aberration and moisture content in bamboo shoot was 150, 1.8, 0.0012, and the root mean square error of calibration set（RMSEC） was 413, 2.3, 0.0042, and the correlation coefficients was 0.985, 0.940, 0.988 respec- tively. With statistical compare between traditional method and near infrared spectroscopy, t value for texture, chromatic aberration and moisture content was 1.93, 0.68 and 1.329 respectively, which was less than the critical value t0.05=2.131. Therefore, there was no significant difference between them, and near infrared spectroscopy would be a rapid, nonde- structive and simple tool for the simultaneous determination of quality parameters in bamboo shoot.