数字舌图的舌色苔色特征提取和选择  

Feature Extraction and Selection of Tong Color and Fur Color in Digital TCM Tong Image

作　　者：彭明德 彭涛[2] 李明 

机构地区：[1]绵阳市中医院,四川绵阳621000 [2]成都微软技术中心,四川成都610000

出　　处：《中华中医药学刊》2009年第9期1876-1878,共3页Chinese Archives of Traditional Chinese Medicine

基　　金：四川省中医药管理局资助项目(200647)

摘　　要：目的:探讨数字化舌诊舌色、苔色标准建立的思路方法。方法:应用文献的方法对临床采集舌图做HSV颜色空间量化,并探讨其与传统舌诊内容的定量关系,以及通过病例探索与中医病机的相关性。结果:从234例舌图分析后选取出7个舌色、苔色特征。222例临床病例选用SPSS的变量间相似度的距离侧度,所选取的7个舌色、苔色特征对中医病机的表达与传统舌诊对中医病机的表达是一致的。结论:HSV颜色空间十分利于传达中医的舌色、苔色的视觉特点;所选取的7个舌色、苔色特征与中医传统舌诊内容兼容;并且,用象素点之和代表各舌色、苔色面积来量化舌色、苔色特征是可行的,可以用来探讨数字舌诊定量表达中医病机属性的规律。Objective： To explore the method of tong color and fur color＇s standard establishment in digital tong diagnosis. Methods： Turning clinical tong images from RGB color space into HSV color space and color quantized by the way which is narrated in literature, then to explore its quantitative relation with traditional tong diagnosis and its correlation with TCM pathogenesis through the cases. Results ： Selected 7 features of tong color and fur color from analyzing 234 cases of tong image; 222 cases of clinical cases of SPSS correlation analysis results can be seen, the selected 7 features and TCM pathogenesis of expression is the same. Conclusion ： HSV color space is quite beneficial to convey the visual characteristics of the tong color and fur color in TCM. The 7 features are compatible with the content of TCM diagnosis. Moreover, it is feasible that we use the sum of pixel points which represent the area of tong color and fur color and we can use it to explore the quantitative expression of TCM pathogenesis attributes in digital tong diagnosis.

关 键 词：舌诊 舌色 苔色 色彩量化 特征选择 

分 类 号：R2-03[医药卫生—中医学]