基于高光谱数据的乳香产地快速鉴别

Rapid Identification of Frankincense Origin Based on Hyperspectral Data

乳香在中医药中有着较高的药用价值,不同产地乳香的药用价值不同,为实现乳香产地分类的快速无损检测,基于高光谱技术,提出一种快速检测方法,以解决中药流通中乳香产地混杂的问题。以三个产地的乳香样品为研究对象,利用高光谱技术,采用归一化方法对光谱数据进行预处理,对预处理后的光谱数据通过连续投影算法(SPA)提取出20个特征波长,以所提取的特征波长作为输入变量,建立极限学习机(ELM)、支持向量机(SVM)和线性判别分析(LDA)三种分类模型,进行预测判别,并与全谱模型的预测准确率进行比较,寻找一种高效的乳香产地判别模型。结果发现,SPA-ELM、SPA-LDA模型的预测准确率均为100%,两种方法均可实现乳香产地的快速、无损鉴别。研究为乳香产地的快速无损检测分析提供参考。

Frankincense had a high medicinal value in traditional Chinese medicine.The medicinal value of frankincense from different producing areas was different.In order to realize the rapid and non-destructive detection of frankincense origin classification,a rapid detection method was proposed based on hyperspectral imaging technology to solve the problem of mixed origin of frankincense in the circulation of traditional Chinese medicine.Frankincense samples from three producing areas were taken as the research objects,hyperspectral imaging technology was used to preprocess the spectral data by normalization method.20 characteristic wavelengths were extracted from the preprocessed spectral data by successive projections algorithm(SPA).Taking the extracted characteristic wavelengths as input variables,three classification models of extreme learning machine(ELM),support vector machine(SVM)and linear discriminant analysis(LDA)were established for prediction and discrimination.Compared with the prediction accuracy of the full spectrum model,an efficient model for distinguishing the origin of frankincense was found.The results showed that the prediction accuracy of the SPA-ELM and SPA-LDA models were both 100%,and both methods could achieve rapid and non-destructive identification of the origin of frankincense,which provided a reference for the rapid non-destructive testing and analysis of frankincense production area.