基于FTIR离散平稳小波特征提取的SVM应用于中药材紫花地丁与同属植物的鉴别研究

    Recognition Method Research of Violae Herba and Its Sibling Plants Based on FTIR-DSWT and SVM Classification Method

    • 摘要: 目的 探索基于FTIR离散平稳小波变换结合支持向量机(support vector machine,SVM)分类法的中药紫花地丁的质量控制新模式。方法 采用衰减全反射傅里叶变换红外光谱法直接快速测定中药紫花地丁与同属植物多花堇菜和戟叶堇菜的FTIR,运用基于离散平稳小波变换进行特征向量的提取,通过分析比较后选取第4、5层分解层的特征向量用于支持向量机的训练与验证。结果 通过对不同产地的90个样本的验证,紫花地丁与同属植物多花堇菜和戟叶堇菜的识别率达100%。结论 基于FTIR离散平稳小波变换结合支持向量机分类法的中药紫花地丁与同属植物多花堇菜和戟叶堇菜的分类鉴别方法具有非常好的效果。

       

      Abstract: OBJECTIVE To explore quality control pattern of new traditional Chinese medicine Violae Herba based on attenuated total reflection-Fourier transform infrared(FTIR) spectroscopy-discrete stationary wavelet transform(DSWT) combined with support vector machine(SVM). METHODS FTIR of Violae Herba and its sibling plants Viola pseudo-monbeigii Chang and Viola betonicifolia J. E. Smith were obtained directly, quickly and accurately by Fourier transform infrared spectrometer with attenuated total reflection accessory. DSWT was used to extrude local region of FTIR of Violae Herba, Viola pseudo-monbeigii Chang and Viola betonicifolia J. E. Smith. After comparison analysis, the d4 and d5 scale were used to extract the feature vectors, which were used to train and test the SVM. RESULTS According to 90 testing samples of the different production from different places, the recognition rate of Violae Herba and its sibling plants were all 100%. CONCLUSION The experimental results show that it is effective to apply DSWT combined with SVM on the basis of FTIR to identify the Violae Herba and its sibling plants Viola pseudo-monbeigii.Chang and Viola betonicifolia J. E. Smith.

       

    /

    返回文章
    返回