不同产地葛根药材的高光谱结合人工神经网络鉴别
Hyperspectral Combined with Artificial Neural Network to Identify Pueraria Lobata Medicinal Materials from Different Origins
郭毅秦 1焦龙 1娄俊豪 1沈瑞华 1钟汉斌 1熊迅宇1
作者信息
- 1. 西安石油大学 化学化工学院,陕西 西安 710065
- 折叠
摘要
采用高光谱结合人工神经网络(ANN)方法建立了不同产地葛根药材的鉴别方法.采集6种不同产地葛根药材的高光谱数据,使用Savitzky-Golay平滑滤波对原始光谱数据预处理,结合人工神经网络方法建立葛根产地鉴别模型.结果表明,与未经预处理的光谱数据模型准确率相比,Savitzky-Golay平滑滤波后建立的ANN模型识别测试集分类准确率达到99.00％.因此,高光谱技术结合人工神经网络能够实现快速、准确地鉴别葛根产地,是一种很有前景的葛根药材鉴别方法.
Abstract
The hyperspectral combined with artificial neural network(ANN)method was used to establish the identification method of Pueraria lobata from different origins.The hyperspectral data of six kinds of Pueraria lobata medicinal materials from different origins were collected,and the original spectral data were preprocessed by Savitzky-Golay smooth filtering,and the origin identification model of Pueraria lobata was es-tablished by artificial neural network method.The results show that compared with the accuracy of the unpreprocessed spectral data model,the classification accuracy of the ANN model established after Savitzky-Golay smoothing filtering reaches 99.00％.The results showed that hyper-spectral technology combined with artificial neural network could quickly and accurately identify the origin of Pueraria lobata,which was a promising method for identifying Pueraria lobata.
关键词
葛根/高光谱/人工神经网络/产地鉴别Key words
Pueraria/Hyperspectral/Artificial Neural Networks/Origin Identification引用本文复制引用
基金项目
国家自然科学基金项目(211723003)
陕西省教育厅青年创新团队建设科研计划项目(21JP097)
陕西省教育厅青年创新团队建设科研计划项目(22JP064)
大学生创新创业训练计划项目(202210700010)
川庆钻探公司-西安石油大学致密油气藏勘探开发研究中心科技项目(CQXA-2023-05)
西安石油大学科研创新团队(2019QNKYCXTD17)
出版年
2024
国家科技期刊平台
NSTL
万方数据