摘要
该文基于红外光谱数据研究中药材产地鉴别问题,首先利用离散小波变换处理原始信号,扩大不同产地光谱图差异,精准筛选出特征波段,然后使用神经网络和判别分析鉴别药材产地。由仿真结果可知,神经网络识别准确率为62%,效果不太理想,而判别分析识别准确率高达89.9%,相较前种方法有较大提升,所以使用判别分析可以快速实现中药材产地鉴别。
In this paper,the origin identification of traditional Chinese medicine is studied based on infrared spectral data.firstly,discrete wavelet transform is used to process the original signal,enlarge the difference of spectral map from different areas,and accurately screen the characteristic band.then neural network and discriminant analysis are used to identify the origin of medicinal materials.The simulation results show that the recognition accuracy of neural network is 62%,and the effect is not ideal,while the recognition accuracy of discriminant analysis is as high as 89.9%,which is much higher than that of the previous method.Therefore,the use of discriminant analysis can quickly realize the identification of the origin of traditional Chinese medicine.
出处
《科技创新与应用》
2023年第17期5-8,共4页
Technology Innovation and Application
基金
安徽省高校自然科学重点研究项目(KJ2021A1523)
安徽省高校自然科学重点研究项目(KJ2020A1107)
安徽省职业教育提质培优行动计划(2020tzpy46-4)。
关键词
小波变换
神经网络
判别分析
红外光谱数据
产地鉴别
wavelet transform
neural network
discriminant analysis
infrared spectral data
origin identification
