基于卷积神经网络的大黄饮片的识别研究  

Identification Research of Rhubarb Decoction Pieces Based on Convolutional Neural Network

作　　者：张怡昕 刘明月 郑乐兵 常皓林 高志宇[1] ZHANG Yixin;LIU Mingyue;ZHENG Lebing;CHANG Haolin;GAO Zhiyu(School of Information Technology,Henan University of Chinese Medicine,Zhengzhou 450046,China)

机构地区：[1]河南中医药大学信息技术学院,河南郑州450046

出　　处：《现代信息科技》2021年第3期72-74,78,共4页Modern Information Technology

基　　金：河南省大学生创新创业训练计划项目(S202010471050);河南中医药大学科研苗圃工程项目(MP2020-95);河南中医药大学大学生创新学习项目(CXXM2020050)。

摘　　要：该文从大黄纹理特征入手研究了卷积神经网络技术在药材识别中的应用。首先介绍了卷积神经网络、研究所需材料及实验设备,然后深入解读了本研究方法技术路线,最后采用卷积神经网络VGG-16构建大黄饮片识别模型。实验结果表明,训练的VGG-16对测试图片集的识别及分类具有较好的效果。从客观性和稳定性来看,采用计算机视觉技术可十分便捷而准确地对市面上的药用大黄与真品药用大黄进行特征匹配,以及对其进行药物鉴别。Based on the texture features of rhubarb,this paper studies the application of convolution neural network technology in the identification of medicinal materials.First,the convolutional neural network,the materials and experimental equipment needed for the research are introduced,and then the method technical route of this research is deeply interpreted.Finally,the convolutional neural network VGG-16 is used to construct the recognition model of rhubarb decoction pieces.The experimental results show that the trained VGG-16 has achieved better effects in the recognition and classification of the test picture set.In terms of objectivity and stability,it is very convenient and accurate to use computer vision technology to match the characteristics of medicinal rhubarb and genuine rhubarb on the market,and to identify the drugs.

关 键 词：卷积神经网络 纹理 VGG-16 大黄 计算机视觉 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] R282[自动化与计算机技术—控制科学与工程]