According to the traditional morphological classification, the Houbo quality of traditional chinese medicine is divided into recommended products and non-recommended products; Discrete the chromatography data of the Houbo which obtained under the condition of standard test and also make the data mining. Obtaining the great peaks of linear independent vectors and obtaining every clustering centre data of Houbo by the fuzzy and Defuzzification methods. Make Houbo criterion and sample data as weibull distribution to calculate similar. This research combines the plant morphology, chemistry, statistics and calculating technology to establish the pattern identification methods of quality and have good attempt to Chinese traditional machine quality methods. The research results provide the basis for quality standard revision of Houbo.