Abstract: Tongue analysis is an essential component of the Four Diagnoses.The efficacy of existing method is not so satisfying when the tongue color samples are overlapped and uncertain.In this paper,we introduce Rough Set theory to tongue color analysis to deal with vagueness and uncertainty cases,experimental result shows the promising effectiveness.

Key words: tongue characterization, rough set, decision table, reduction

摘要： 舌诊是中医四诊中望诊的重要内容,已有的舌色苔色分析方法在识别颜色特征空间互相重叠和不精确样本时效果欠佳。利用粗糙集在处理不完备信息方面的优势,首次将粗糙集理论用于舌色苔色分析中。采用量化的方法以减少颜色属性值;提出了一种按独立属性个数递增顺序进行约简的方法以及一种对待测样本在上、下近似集分别打分的综合分类识别算法,实验结果证明了该文方法的有效性。

关键词: 舌诊客观化, 粗糙集, 决策表, 约简