北京生物医学工程

基于舌图像深度特征融合的中医体质分类方法研究

Preliminary study on traditional Chinese medicine constitution classification method based on tongue image feature fusion

作者：周浩胡广芹张新峰

单位：北京工业大学信息学部（北京 100124）

关键词：卷积神经网络；特征融合；舌图像；体质类型分类；中医

分类号： R318.04

出版年·卷·期（页码）：2020·39·3（221-226）

摘要：

目的在基于舌图像的中医体质类型分类中，舌图像的类间距小，传统手工特征提取时存在底层特征不能够充分表达舌图像所包含的全部信息等问题。因此本文提出一种基于深度网络特征层融合的体质类型分类方法，以提高体质类型分类的准确率。方法通过比较不同网络模型对舌图像的分类表现，及对不同网络层的特征表达能力的分析，选取将浅层特征与高层语义特征进行融合的方法。该深度特征融合方法基于Alexnet网络进行改进，依据误差权重，对各层特征进行融合。并采用983张舌图像，对气虚质、痰湿质和湿热质三种体质类型的分类问题进行仿真实验。结果相比传统特征提取与原始深度网络，本文方法的准确率由传统分类方法的54.3%提高到了77%。结论基于深度特征融合的方法将浅层特征与深度特征融合，充分表达了图像的语义信息，对中医辅助辨识、临床、教学和科研具有极其重要的研究意义。

Objective The traditional manual feature extraction is mainly based on the bottom features, which can not adequately express all the information contained in the tongue image. In order to improve the accuracy of physique classification, a method of physique classification based on feature layer fusion of deep network is proposed in this paper. Methods Different network models have different classification performance for tongue images, and different network layers have different feature description capabilities. This method fuses shallow features with high-level semantic features, avoides the interference of human factors in traditional special extraction. The method of depth feature fusion is improved on Alexnet network. According to the error weight, the features of each layer are fused. Based on 983 tongue image data sets, this paper discusses the classification of three constitutional types: Qi deficiency, phlegm-dampness and damp-heat. Results The simulation results show that compared with the traditional feature extraction and the original depth network, the accuracy of this method is improved from 54.3% to 77%. Conclusion The method based on depth feature fusion is proposed to fuse shallow features with depth features, which fully expresses the semantic information of images. It is of great significance to the research of assistant identification, clinical, teaching and scientific research of traditional Chinese medicine.

参考文献：

[1]李晓彦.治未病理念在体检亚健康人群健康指导中的应用效果探析[J].护理研究,2018,32(20):3309-3310. Li XY.Probe into efect of concept of zhiweibing for health examination of sub-health crowds[J].Chinese Nursing Resarch ,2018,32(20):3309-3310.

[2]俞若熙,王琦,王济,等.体质辨识应用研究现状分析[J].中国中医药信息杂志,2013,20(2):107-109.

[3]洪乐,叶双林,刘婷婷,等.基于舌象分析的中医体质辨识系统的研究思路[J].中国中医药现代远程教育,2016,14(2):34-36. Hong L,Ye SL,Liu TT, et al.Research on constitution of traditional Chinese medicine identification system based on tongue manifestation[J].Chinese Medicine Modern Distance Education Of China,2016,14(2):34-36.

[4]王琦.9种基本中医体质类型的分类及其诊断表述依据.北京中医药大学学报,2005,28(4):1-8. Wang Q,Classification and diagnosis basis of nine basic constitutions in Chinese medicine[J].Journal of Beijing University of Traditional Chinese Medicine,2005,28(4):1-8.

[5]李博聪,黄庆梅,陈松鹤,等.基于CIELAB空间的中医舌色分析方法[J].世界科学技术中医药现代化,2007,9(3):28-32. Li BC,Huang QM,Chen SH, et al. CIELAB color space based classification of tongue colors[J].World Science and Technology Modernization of Traditional Chinese Medicine and Materia Medical,2007,9(3):28-32.

[6]张新峰,沈兰荪,刘垚巍,等.基于粗糙集与SVM概率输出的中医舌象特征融合方法研究[J].世界科学技术：中医药现代化,2007,9(5):122-128. Zhang XF,Shen LS,Liu YW,et al.Ａ Study of tongue manifestation feature fusion using rough set and SVM probability[J].World Science and Technology Modernization of Traditional Chinese Medicine and Materia Medical,2007,9(5):122-128.

[7]胡继礼,阚红星.基于卷积神经网络的舌象分类[J].安庆师范大学学报(自然科学版),2018(4):44-49. Hu JL,Kan XH.Tongue classification based on convolutional neural network[J].Journal of Anqing Normal University(Natural Science Edition),2018(4):44-49.

[8]杨晶东,张朋.基于迁移学习的全连接神经网络舌象分类方法[J].第二军医大学学报,2018,39(8):897-902. Yang JD,Zhang P. Tongue image classification method based on transfer learning and fully connected neural network[J]. Academic Journal of Second Military Medical University, 2018,39(8):897-902.

[9]Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems. Red Hook, NY:Curran Associates Inc,2012:1097-1105.

[10]Szegedy C , Liu N W , Jia N Y , et al. Going deeper with convolutions[C]// 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston：IEEE Computer Society, 2015.

[11]He K,Zhang X, Ren S,et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. Las Vegas：IEEE, 2016: 770-778.

[12]韩立博, 胡广芹, 张新峰, 等.基于直方图均衡化的伽马校正和K-means聚类的舌象苔质分离方法[J]. 北京生物医学工程, 2019,38(1):1-6. Han LB,Hu GQ,Zhang XF,et al. Separation method of tongue coating and body of tongue image based on histogram equalization and gamma correction and K-means clustering[J].Beijing Biomedical Engineering,2019,38(1):1-6.

[13]刘斌, 胡广芹, 张新峰, 等. 一种改进的中医舌苔厚度自动描述方法[J]. 北京生物医学工程, 2018, 37(2): 157-163. Liu B,Hu GQ,Zhang XF,,et al. An improved automatic description method of tongue coating thickness in Chinese medicine[J] .Beijing Biomedical Engineering,2018, 37(2): 157-163.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录