Identifying Constitution Of Traditional Chinese Medicine Via Tongue Image Based On Deep Neural Network

Tongue diagnosis is a unique and important part of traditional Chinese medicine.It is an important method observing the tongue color and shape changes in diagnosis.In this paper,I use modern computer technology to analyze the tongue image and combine the result of analysis with TCM theory and expert experience.In addition,I adopt the constitution identification method to judge the TCM Constitution quickly and intelligently.It is significant to the inheritance,innovation and promotion of traditional Chinese medicine.Deep learning is a popular approach to machine learning,and it has made gratifying achievements in many fields such as image recognition,object detection,natural language processing and so on.This paper uses deep learning,applying the deep neural network model to tongue detection,tongue calibration and constitution of TCM identification.It aims at solving current problems in the research of tongue analysis which requires a specific environment or professional equipment.With the use of deep learning and large amounts of data,any smart mobile device can be used in tongue analysis.Because there is not enough and high-quality tongue image in tongue analysis,this paper designs a data collection system which uses smart mobile devices to collect tongue images in hospital.In this paper,I build the database for tongue detection,tongue calibration and constitution of TCM identification and implement these with deep neural network.Finally,I design a system that can do tongue analysis with smart mobile devices.