Face diagnosis is one of the four diagnostic methods in Traditional chinese medicine（TCM）. The morbidity of the organs can be revealed from the facial complexion. Due to the ambiguity of face diagnosis in TCM,Fuzzy support vector machine（FSVM） is utilized to remove the influence of outliers. The facial complexion recognition in TCM based on FSVM is proposed in this paper,which includes the following steps: 1) The facial cheek region are segmented as skin blocks; 2) The color feature in Lab color space is extracted from skin block to represent facial complexion characteristic; 3) The fuzzy membership is calculated to obtain the membership of the training samples, in which the final membership of training samples is determined by combining membership calculation based on distance with the facial complexion recognition based on color modeling; 4) The FSVM is built to classify facial complexion characteristic. Experimental results show that proposed facial complexion recognition has better antiinterference performance as well as a higher recognition rate up to 82%.
展开▼
