为探究一种快速、 可靠的肉苁蓉属中药材检测方法, 实验采用荧光光谱成像技术结合模式识别方法对肉苁蓉属三种中药材: 荒漠肉苁蓉、 管花肉苁蓉和沙苁蓉进行鉴别研究。 实验中发现肉苁蓉样品存在较显著的荧光特性, 采集来自不同产地、 不同批次以及不同超市购买的三种肉苁蓉属药材的40个样品的荧光光谱图像, 对图像进行去噪、 二值化处理后, 根据光谱立方体绘制每个样本的光谱曲线, 将所得450～680 nm波段范围内的光谱数据作为鉴别分析的研究对象, 应用主成分分析法(PCA)对三种肉苁蓉的光谱数据进行降维处理, 再结合Fisher判别方法对三种肉苁蓉进行鉴别。 分别比较多元散射校正(MSC)、 标准正态变量校正变换(SNV)以及一阶微分(FD)三种数据预处理方法对鉴别模型的影响, 并根据主成分的累积贡献率和主成分因子数对判别模型效果的影响对主成分因子数进行优化。 分析结果表明: 一阶微分预处理后提取前四个主成分进行Fisher判别的鉴别效果最佳, PCA结合Fisher判别建立肉苁蓉属三种药材的判别模型原始判别的准确率达到100%, 交叉验证的准确率达到95%。 由此可见, 利用荧光光谱成像技术结合主成分分析及Fisher判别对肉苁蓉属三种药材的鉴别分析是可行的, 而且具有操作简便、 快速、 可靠等优点。

In order to explore rapid reliable Hebra cistanche detection methods, identification of 3 different sources of Hebra cistanche: cistanche deserticola, cistanche tubulosa, sand rossia is studied via fluorescent spectral imaging technology combined with pattern recognition. It is found in experiment that cistanche samples have obvious fluorescence properties. Forty fluorescence spectral images of 3 different sources of Hebra cistanche samples are collected through fluorescent spectral imaging system. After carrying on denoising and binarization processing to these images, the spectral curves of each sample was drawn according to the spectral cube. The obtained spectra data in the 450～680 nm wavelength range is regarded as the study object of discriminant analysis. Then, principal component analysis (PCA) is applied to reduce the dimension of spectroscopic data of the three kinds of cistanche and fisher distinction is used in combination to classify them; During the experiment were compared the effects of three methods of data preprocessing on the model: multiplicative scatter correction (MSC), standard normal variable correction (SNV) and first-order differential (FD) and then according to the cumulative contribution rate of the principal component and the effect of number of factors on the discriminant model to optimize the number of principal components factor. The results showed that: identification of the best after the first derivative pretreatment then the first four principal components is extracted to carry on fisher discriminant, discriminant model of 3 different sources of Hebra cistanche is set up through PCA combined with fisher discriminant the precision of original discrimination is 100%, recognition rate of the cross validation is 95%. It was thus shown that the fluorescent spectral imaging technology combined with principal components analysis and fisher distinction can be used for the identification study of 3 different sources of Hebra cistanche and has the advantages of easy operation, speediness, reliability.