雷达图表示与相似度计算相结合的麻黄品种鉴别方法

发布时间：2019-02-17 20:40

【摘要】：将指纹图谱技术用于中药麻黄的鉴别,提出了指纹图谱雷达图表示与相似度计算相结合的麻黄品种鉴别分析方法。该方法首先对麻黄样品的HPLC指纹图谱的峰面积分别进行相似度计算和雷达图表示,然后对雷达图进行重心图特征提取,最终得到用于分类比较的雷达图特征,最后结合雷达图特征与相似度特征对麻黄品种进行鉴别。实验结果显示该方法鉴别准确率明显高于只采用相似度模式特征的鉴别,且综合特征的最高分类准确率达98.25%,表明雷达图表示与相似度特征相结合的方法是一种可行的麻黄品种鉴别方法。
[Abstract]:The fingerprint technique was applied to the identification of ephedra, and a method for identification and analysis of Ephedra ephedra was put forward in this paper, which was based on the combination of fingerprint radar image representation and similarity calculation. In this method, the peak area of HPLC fingerprint of ephedra samples is calculated by similarity calculation and radar image representation, then the barycenter feature of radar image is extracted, and finally the radar image features used for classification and comparison are obtained. Finally, combined with radar image features and similarity features to identify the varieties of ephedra. The experimental results show that the accuracy of this method is significantly higher than that of only using similarity pattern features, and the highest classification accuracy of the comprehensive features is 98.2525. It shows that the combination of radar map representation and similarity feature is a feasible method for identification of Ephedra equisetifolia.
【作者单位】：河北省测试计量技术及仪器重点实验室;秦皇岛职业技术学院旅游系;燕山大学电气工程学院;中国中医科学院针灸研究所;
【基金】：中国博士后科学基金(2012M510722) 中国中医科学研究院项目(ZZ12001) 燕山大学博士基金(B692) 秦皇岛市科学技术研究与发展计划(201001A119)资助项目
【分类号】：TN957.52

【参考文献】