A New Approach To The Tongue-Image Segmentation And Moistening Analysis Based On Image Processing

As one of the quintessence of Chinese traditional culture,traditional Chinese medicine(TCM)is more and more popular because of its ideology.Compared with traditional Chinese medicine,modern Chinese medicine pays more attention to the objective and intelligent.Since tongue diagnosis is an important part in the "inspection" of TCM,segmentation and analysis of tongue are important contents in the modernization process of TCM.This paper focuses on the tongue-image segmentation and the recognition of tongue coating.First of all,we used the gray integral projection and improved median filter for pre-segmentation,and then use the feature vectors for clustering based on kernel fuzzy cluster means.Secondly,in HSV color space,the body and coating of tongue were separated,and a number of features were extracted and analysed.Combined with TCM theory,these characteristics were associated with physical signs.Among them,the characteristic of tongue moistening was analysed by using binary optical reflection model and moistening coefficient.Finally,in the field of tongue-image acquisition,in order to get the clear and complete tongue-image,we designed a new type of tongue-image acquisition device according to the mechanical structure and function demand perspective.At the same time,we developed the portable tongue diagnosis system based on mobile platform which achieved the main function of the tongue diagnosis instrument.The results of experiments show that the cluster algorithm can complete the segmentation of tongue-image with high accuracy.After the separation of tongue body and coating,the moistening of tongue coating can be accurately identified and quantified.The portable tongue diagnosis system with the tongue-image acquisition device,can complete the tongue-image acquisition,processing,analysis and diagnosis.It's proved to be practical,comfortable and efficient.