A Study Of Tongue Segmentation From Image And Tongue Recognition And Classification Algorithm

Tongue diagnosis of TCM(traditional Chinese medicine)is the main content of "look" diagnosis and a traditional diagnostic method which is the most characteristics of traditional Chinese medicine.Tongue also reflects the most sensitive indicators of the physiological function and pathological changes,in the process of diagnosis and treatment of TCM has important application value.With the improvement of modern science and technology and the continuous development of image processing technology,image processing technology have been widely used in various fields.The combination of tongue diagnosis and image processing technology can realize the automatic diagnosis of tongue,which can avoid the defects such as non-quantification,uncertainty and misdiagnosis and so on,and the automatic diagnosis of tongue has been an emerging research.This paper mainly researches the tongue segmentation technology and the classification and recognition technology of the tongue.The contents can be summarized as follows:First of all,this paper realizes the rough segmentation to get a part of the picture which includes tongue through finding the position of the tongue and cutting out this part manually.Then,for the question of tongue auto-segmentation about the lack of contour and color information,this paper comes up with an arithmetic of inner product projection segmentation to enhance the robustness of image segmentation.At the same time,in order to supplement the edge information of tongue and the color information that segmentation dismissed,this paper comes up with the color clustering method to enhance the completeness of tongue information.The experimental results show that the method proposed in this paper has a significant improvement compared with the traditional segmentation method in the speed and the integrity of the tongue segmentation.Finally,according to the diversity of tongue image and the nonlinear characteristics of tongue body,this paper designs a nonlinear classifier of multi class based on the principle of SVM.It is able to quickly and effectively determine the type of tongue on the basis of the principal component analysis to reduce the dimension of the tongue feature data.The experimental results show that the PCA can obviously reduce the dimension of data,which can diminish the amount of computation and diminish the time of classification.In addition,the accuracy of the classifier designed in this paper can reach more than 90%.In this paper,a series of tongue diagnosis methods are proposed,which can realize the complete segmentation of tongue and the recognition and classification of tongue.These methods have good adaptability and accuracy,and have a certain reference value for the existing automatic tongue diagnosis system.It is of great significance for the research and development of tongue diagnosis modernization.