浏览全部资源
扫码关注微信
1.扬州市职业大学 医学院,江苏 扬州 225000
2.南京中医药大学 药学院,南京 210000
* 陈雁,硕士,从事中药鉴定、深度学习方面的研究,E-mail:chenyan1987_2010@163.com
纸质出版日期:2021-08-05,
网络出版日期:2021-04-30,
收稿日期:2021-01-30,
扫 描 看 全 文
陈雁,邹立思.基于BMFnet-WGAN的中药饮片智能甄别[J].中国实验方剂学杂志,2021,27(15):107-114.
CHEN Yan,ZOU Li-si.Intelligent Screening of Pieces of Chinese Medicine Based on BMFnet-WGAN[J].Chinese Journal of Experimental Traditional Medical Formulae,2021,27(15):107-114.
陈雁,邹立思.基于BMFnet-WGAN的中药饮片智能甄别[J].中国实验方剂学杂志,2021,27(15):107-114. DOI: 10.13422/j.cnki.syfjx.20210819.
CHEN Yan,ZOU Li-si.Intelligent Screening of Pieces of Chinese Medicine Based on BMFnet-WGAN[J].Chinese Journal of Experimental Traditional Medical Formulae,2021,27(15):107-114. DOI: 10.13422/j.cnki.syfjx.20210819.
目的
2
为适应现代化饮片甄别的需求,克服传统人工经验方法主观性强而效率低的问题,探究机器视觉与深度学习方法在中药饮片智能甄别领域的可行性具有重要的研究意义。
方法
2
构建包含60种11 125张饮片的图像集,设计高低频特征学习的网络架构,即采用平行卷积网络获得低频特征和多尺度深度卷积核获得高频特征,并利用语义描述网络实现具备泛化能力的特征学习模式。研究将Wasserstein间距引入博弈生成对抗模型完成饮片甄别,在生成和判别网络中增加条件参数,使网络训练更可靠,同时提升识别精准度。
结果
2
实验表明当训练样本与测试样本占比大于6∶4时,饮片识别准确度较为稳定;该文方法针对不同状态和环境下捕获的饮片图像,平均甄别精准度最能达85.9%,稳定度高,识别效果显著优于VGG-Net和AlexNet方法。
结论
2
该文方法能够获得丰富而典型的饮片特征,所引入的沃瑟斯坦生成对抗神经网络(WGAN)模型和Wasserstein间距可使网络训练更可靠;所完成各种复杂环境下的中药饮片智能甄别准确度、鲁棒性和批量化效果好,为饮片分拣与质量量化甄别给出技术支撑。
Objective
2
To explore the feasibility of machine vision and deep learning methods in intelligent screening of pieces of Chinese medicine, so as to meet the needs of modern screening of pieces of Chinese medicine and overcome the problems of strong subjectivity and low efficiency in traditional screening based on manual experience.
Method
2
An image set containing 11 125 images for 60 kinds of pieces of Chinese medicine was constructed, and the network architectures for high- and low-frequency feature learning were designed. Specifically, the parallel convolutional network was employed to obtain the low frequency feature and the deep multi-scale convolutional neural network to uncover the high-frequency feature. The semantic network was used to realize the feature learning mode with generalization ability. In this study, Wasserstein distance was introduced into the generative adversarial networks (GANs) to complete the screening of pieces of Chinese medicine, and the conditional parameters were added to the generation and discrimination networks to make the network training more reliable and improve the accuracy of identification.
Result
2
The experiment results showed that when the ratio of training samples to test samples was greater than 6∶4, the identification accuracy of pieces of Chinese medicine was relatively stable. The identification accuracy of images captured in different states and environments by bi-view multi-feature network Wasserstein generative adversarial network (BMFnet-WGAN) reached up to 85.9% on average and the stability was high, demonstrating that BMFnet-WGAN was superior to VGG-Net and AlexNet.
Conclusion
2
The BMFnet-WGAN method enables the revealing of rich and typical characteristics of decoction pieces and the introduced WGAN model and Wasserstein distance make the network training more reliable. The resulting accuracy, robustness, and batch effects in the intelligent screening of pieces of Chinese medicine were good, which has provided the technical support for the sorting and quantitative quality screening of pieces of Chinese medicine.
中药饮片特征学习语义描述Wasserstein
traditional Chinese medicine (TCM)piecesfeature learningsemantic descriptionWasserstein
刘珊,苏泽琦,刘逍遥,等.基于PubMed和Web of Science数据库对中医药治疗慢性萎缩性胃炎现状的文献分析[J].中国实验方剂学杂志,2021,27(6):149-158.
谢景,镇兰萍,任广喜,等.早臺对白花前胡根部解剖结构和香豆素类成分的影响[J].中国中药杂志,2020,45(20):4861-4866.
雷林,肖锟钰,瞿领航,等.基于颜色量化技术的谷芽及其炮制品外观性状判别标准的建立[J].时珍国医国药,2020,31(8):1877-1881.
陈仕妍,卢文彪,王凤梅.基于颜色匹配模板的中药饮片图像识别[J].中国实验方剂学杂志,2020,26(6):158-162.
谢树莹,周明,周金海.基于OpenCV的中药饮片“辨色”识别研究[J].时珍国医国药,2018,29(2):510-512.
谭超群,温川飙,吴纯洁.基于图像处理技术的中药饮片识别研究[J].时珍国医国药,2018,29(7):1706-1709.
陶欧,林兆洲,张宪宝,等.基于饮片切面图像纹理特征参数的中药辨识模型研究[J].世界科学技术—中医药现代化,2014,16(12):2558-2562.
吴冲,谭超群,黄永亮,等.基于深度学习算法的川贝母、山楂及半夏饮片的智能鉴别[J].中国实验方剂学杂志,2020,26(21):195-201.
李佳园,魏晓嘉,万国慧,等.“辨状论质”的历史沿革与现代研究进展[J].中国实验方剂学杂志,2021,27(6),189-196.
赵玉姣,彭华胜.历代本草何首乌图考及其“辨状论质”观[J].中国中药杂志,2020,45(8):1960-1967.
ZHANG X,WU H W,YU X K,et al.Determination of bitterness of andrographis herba based on electronic tongue technology and discovery of the key compounds of bitter substances[J].Molecules,2018,23(12):3362.
张天翼,郭成杰,郭宁,等.药效相关的中药感官指标测定方法的研究进展[J].中草药,2021,52(2):594-602.
刘天睿,金艳,孟虎彪,等.论中药“辨状论质”之辨色泽与品质评价的生物学内涵研究[J].中国中药杂志,2020,45(19):4545-4554.
刘瑞新,郝小佳,张慧杰,等.基于电子眼技术的中药川贝母真伪及规格的快速辨识研究[J].中国中药杂志,2020,45(14):3441-3451.
万超,于定荣,刘颖,等. 醋延胡索饮片颜色与其内在质量的相关性分析[J].中国实验方剂学杂志,2019,25(12):145-150.
杨丽,陈鸿平,李雪莲,等.不同“变色”程度枸杞子外观颜色表征与内在色素类成分变化相关性[J].中国实验方剂学杂志,2015,21(8):47-50.
甄臻,王杨,魏海峰,等.基于颜色变化的麸炒山药质量标准及炮制工艺探究[J].中成药,2021,43(3):816-819.
焦美钰,佟月,吴浩善,等.南、北五味子“辨色论质”机制及其定性识别模型研究[J].中国药房,2020,31(24):3007-3012.
陈金素,向金莲,杨江媛.六个不同地区的中药炮制特色技术浅析[J].华西药学杂志,2021,36(1):95-98.
洪智慧,杜伟锋,李小宁,等.中药材产地趁鲜加工的可行性及相关建议[J].中华中医药杂志,2021,36(01):80-85.
张联.中国药典“炮制通则”中药炮制方法分类探讨[J].中国现代应用药学,2020,37(18):2287-2290.
芦柏震,何晓波,章红燕.改进中药材的浸泡、润软化处理提高中药饮片质量[J].中草药,2001,32(8):43.
谢杰.浅谈切制前的软化处理对饮片质量的影响[J].中国现代药物应用,2012,6(18):121.
LI W,YAN F,GAO Y,et al.Research on safety and management countermeasures of experimental teaching based on traditional Chinese medicine[J].WSRJ,2019,5(12):173-177.
FIGUROWSKI D,DWORAK P.Environment mapping algorithm using semantic description and constrained delaunay triangulation[J].Elektron Elektrotech,2019,25(6):4-7.
LIU B,ZHANG Q,GE X,et al.CVaR-based approximations of Wasserstein distributionally robust chance constraints with application to process scheduling[J].Ind Eng Chem Res,2020,59(20):9562-9574.
DIPLAROS A,VLASSIS N,GEVERS T.A Spatially constrained generative Model and an EM algorithm for image segmentation[J].Ieee Trans Neural Networ,2007,18(3):798-808.
陶益,陈林,江恩赐,等.人工智能和工业4.0视域下高光谱成像技术融合深度学习方法在中药领域中的应用与展望[J].中国中药杂志,2020,45(22):5438-5442.
0
浏览量
17
下载量
3
CSCD
关联资源
相关文章
相关作者
相关机构