中文摘要：

舌诊是中医诊法最重要的特色之一，舌诊客观化研究有助于舌诊的继承和发展。根据舌象的特点，分析并处理了舌象的色度直方图，提出了自动确定颜色类别数和初始聚类中心的算法，并在改进模糊聚类算法的基础上提出了一种彩色舌图像舌质舌苔分离方法。实验结果表明，与标准模糊聚类算法相比，该算法极大地加快了聚类迭代速度，减少了系统运算时间；相对于阈值分割法，能够获得更加符合中医要求的苔质分离效果。

英文摘要：

Tongue inspection is one of most important characteristics of TCM diagnostic methods. The objectivity research of tongue inspection would improve the traditional medical method. According to tongue image feature, through analysis and process of hue histogram of tongue image, proposed a novel approach to automatically determine color classed and initial clustering centers. Modified the fuzzy C-means algorithm and presented a novel algorithm of tongue body and tongue coating sepa- ration. The experimental result indicates that this algorithm speeds up the clustering iteration enormously and reduces the sys- tem operation expenses comparing with standard FCM. The tongue body and tongue coating separation effect conforms to Chi- nese medicine request better than effect by threshold value segmentation method.