Study On Diagnosis System Of Traditional Chinese Medical Tongue Image Based On Digital Image Processing

Objective:Tongue diagnosis is a part of TCM Inspection. It is an important symbol of traditional experience and Characteristics of Chinese Medical four diagnosis and also plays an important role in traditional diagnosis, which is the diagnosis and treatment basis for Various Diseases in clinic.However the diagnosis is obtained by the methods of TCM doctor's eye observation during treatment. It leads to an inevitably artificial, objective result and is difficult to make exact index. Dissertation combines the modern Image Processing with clinical experience of TCM expert. The researcher builds up an automatic digital tongue diagnosis system to increase the relative theoretical level and clinical diagnostic ability in order to implement the quantization, Objectification and Standardization of tongue diagnosis.Methods:Tongue Image apparatus is constituted by tongue image acquisition system and tongue image characteristic processing system. Acquisition system is mainly made up with light sources, lighting environment, digital camera and some other hardware. Tongue Image characteristic processing system is constituted by the soft of color correction, Tongue Segmentation, tongue body and tongue coating separation, tongue body and tongue coating color recognition, and diagnosis Technology. Dissertation uses the LED as apparatus light source and integrating spheres as illuminant geometry condition. Obtain the tongue image by digital camera. We separate the tongue body image by Greedy algorithm based on classical Snake model and then segment tongue body and tongue coating by the method of dynamic threshold. In the HSV color space, the color of tongue and tongue coating are identified and diagnosed based on statistical pattern recognition theory.Results:It can reduce the volume of apparatus and increase the Electrical Security using the LED as mainly light source. It basically solves the difficulties of differentiate between tongue coating and saliva causing by reflection, which increase the Uniformity of light, on the integrating spheres as illuminant geometry condition. And then the color of tongue and tongue coating are identified and diagnosed when the tongue body, tongue coating and tongue proper separation have been finished. Conclusion:Most images tongue body separation has favorable effects and there are always some errors owing to objective factors. In the study of separation of tongue proper and tongue coating by the method of dynamic threshold,the result had perfect separation efficiency when the tongue coating displays significantly and boundary distinct on the contrary to the tongue coating distribution scattered and boundary fuzzy.