Research And Application Of Tongue Body Image Segmentation And Feature Extraction Technology

Tongue diagnosis is an important part of the see diagnosis in Traditional ChineseMedicine(TCM).Is a necessary component in clinical diagnosis. It has the advantage that noinjury, no pain and easy to use Famous doctors, who lived in ancient society or live in modernsociety, are good at tongue diagnosis. Chinese medicine think that tongue image can objectivelyreflect the human body in a variety of physiological and pathological changes, showing that atthe time of reaction and the functional state of human body; At the same time, the modernclinical practice also shows that, in some cases, tongue diagnosis can found the lesions whichcan’t be check out through modern medical equipmentThe doctor through direct and qualitative observation to get the diagnosis in TCM .Thisresult often be influence by the diagnosis time, diagnosis place and the doctor. This gives thefurther development of tongue diagnosis brought a very big difficulties. Therefore conducts theresearch through the information technology to objective of tongue diagnosis, For TCMstandardization and TCM scientific research teaching method modernization has importantPractical significance.In 1980’s, Chinese and foreign specialist began with tongue diagnosis standardization. Atthat time, they emphasized on color recognition by measuring and comparing the tongue.Recently, with the image segmentation, texture analysis, pattern recognition and other newtechnology, new method is introduced, so that the tongue diagnosis objective research has madegreat progress.First the tongue diagnosis in Chinese is introduced briefly in the paper, and points out theexisting problems, and then the tongue objective and research difficulties are discussed.Because the digital image processing technology in objective of tongue diagnosis played animportant role, so more detailed introduction of digital image processing technology.The work done by this paper has the following three aspects: First of all, The correctsegmentation of the body of tongue is a difficult problem to the information on diagnosis by thefeature of tongue in TCM . using traditional image segmentation methods are very difficult to obtain better segmentation result, therefore, it is necessary to targeted on the tongue areasegmentation method research. Through researching and Experimenting summary the Imagesegmentation and feature extraction，discovered Where is the difficult for the segmentation ofbody of tongue is concluded And an improved algorithm is advanced in the paper it Across thespace component extraction and integration then using the Snake model updating of tongueimage segmentation method, the algorithm in the acquisition of tongue image prior knowledge,comprehensive use of decorrelation operation, Otsu law and Snake model .Technology to realizethe tongue area segmentation; And then, we talk about extraction and presentation of tonguefeatures based on geometry, color and texture. In order to Automatic recognition the tongueimage. Finally, the Image segmentation of image of body of tongue and feature extraction issummarized and the related research in the future is prospected.