粮食芽变初始于内部胚芽部分, 萌芽早期活动发现困难, 成为制约粮食储备安全的一个瓶颈。 粮食芽变过程主要是将淀粉转化为用于生长的麦芽糖。 采用太赫兹(Terahertz, THz)成像技术, 研究不同浓度的小麦麦芽糖混合物的图像特征, 在图像中样品区域各个像素点提取出一条THz光谱, 并求平均, 得到一条平均光谱。 通过主成分分析方法(principal component analysis, PCA)前5个主成分的累计方差贡献率达到98%以上, 用前5个主成分的得分值矩阵代替原始光谱数据, 建立基于径向基核函数的支持向量机(support vector machine, SVM)定量分析模型。 对比结果表明, 太赫兹成像技术结合化学计量学方法对小麦麦芽糖含量具有较好的判别效果, 其中SVM模型获得最好的预测结果, 为通过小麦种麦芽糖含量实现对小麦芽变程度的检测, 保障粮食品质, 提升国家储藏科技水平奠定了理论和技术基础。

Grain germination begins in the interior embryo, the early germination is difficult to be detected, which is a bottleneck restricting the safety of grain storage. The biological process of wheat grain germination mainly translates the starch into maltose for growth. In this paper, Terahertz(THz)imaging technology is used to investigate the image characteristics of maltose with different concentration. A THz spectrum is extracted from each pixel of the sample area in the image, and an average spectrum is obtained by averaging. The cumulative variance contribution rate of the first 5 principal components is over 98% according to the Principal Component Analysis(PCA) method, then the support vector machine (SVM) quantitative analysis regression model based on radial basis function is constructed by using the first 5 principal components score matrix. The comparison showed that THz imaging technology combined with chemometrics method is efficient and feasible for identifying maltose of different concentration and also indicated that the SVM obtains the best prediction results, which provides the theoretical basis for the detection of the state of germinated grain, guarantees the quality of stored grain and enhances the level of national storage technology.