The Research Of Extraction Of The Tongue Image's Feature & Analysis Technology Of Healthy Information Based On Traditional Chinese Medicine Inspection

The tongue is regarded as one of the epitomes of viscera and organ of human body in the TCM. The TCM Tongue Inspection plays important role in diagnosing and treating all kinds of diseases because the symptoms of many illnesses can be reflected in the different parts of the tongue through the nerves system, blood vessels and channels. But the traditional tongue diagnoses methods lack stability in reusing because of man-made subjectivity, which is based on the intuitive qualitative observation of the doctors. In order to solve the above problems, the thesis analyzes the tongue images and simulates the process of the TCM inspection with the ways of computer image processing and computer pattern recognition to achieve objective and automatic TCM Tongue Inspection and, meanwhile giving advice to improve health condition.At first, the thesis analyzes the issue of separation of the tongue coat and tongue body. The separation results directly affect the following recognition quality of tongue body color and tongue coat texture, finally, affects the accuracy of the healthy information given to the clients. The thesis comprehensively applies the Optimal Threshold Iterative Algorithm based on HSV Color Space and Separation Algorithm based on Color Features Combination, effectively solving the problem of separation of tongue coat and tongue body.The next issue is recognizing the tongue body color and tongue coat texture which is a very important aspect in feature extraction of tongue image. SVM with concise mathematical form, visual geometric explanation and good generalization ability is widely used in classification issues. The SVM is applied to the features research of tongue image which is processed by the separation of tongue body and tongue coat, which makes perfect results in classification of tongue body and tongue coat.Then, the thesis selects the tip of tongue as the sub-parts analysis starting point, focus on the research of red tip of tongue and petechia. The thesis use the HSI Color Model to research the red tip of tongue and better completes the research by taking full advantage of the color components of the model. The thesis uses the Region Growing Algorithm to allocate the petechia and ecchymosis area, then, distinguishes the acantha and the petechia of the tongue tip according the features of petechia color, finally, gives a better analysis of whether there are petechia and ecchymosis on the tongue tip by taking full advantage of digital image processing method.Once again, according to the potential healthy information reflected by tongue features analyzed in the previous steps, combining the client's present healthy status, the thesis gives appropriate diet guidance and attention in the daily life, in the end, achieving the purpose of integration of features extraction of tongue image and healthy information analysis.At last, the thesis makes a summary of the extraction of tongue image features based on TCM inspection and technology of healthy information analysis, and gives a suggestion on the further research on the issue.