AdaBoost算法在中医舌诊图像分区识别中的研究  

Research of AdaBoost Algorithm in Tongue Images of Traditional Chinese Medicine Partitions Recognition

作　　者：张萌[1] 胡显伟[1] 王元斌[1] 张斌[1] 

机构地区：[1]东北大学计算机应用技术研究所,辽宁沈阳110004

出　　处：《小型微型计算机系统》2008年第6期1149-1153,共5页Journal of Chinese Computer Systems

基　　金：国家科技部"十五"攻关项目"中医辨证规范与疗效评价的研究"(2004BA721A05)资助

摘　　要：舌诊是中医传统诊断的一个重要方式.本文以模式识别、图像处理和中医舌诊理论为研究基础,结合中医诊断规则,提出舌象的分区训练识别方法,并采用一种新的基于集成学习的AdaBoost算法,构建出一套完整的、符合中医诊断体系规范的舌象分类识别算法.通过实验验证,该方法对中医舌诊分类识别是有效的.Tongue diagnosis is an important method of traditional Chinese medicine （TCM） diagnosis.This paper based on pattern recognition,image processing and TCM tongue theories,and combined with TCM diagnostic rules.Then it introduced a training and recognizing method of tongue partitions,and adopted the AdaBoost algorithm that was a new algorithm based on integrated learning. Finally, it constructed a complete tongue diagnosis classification algorithm that accorded with the system standard of TCM diagnosis.Experiments showed that the method is valid to TCM tongue classification.

关 键 词：ADABOOST算法 分区 分类器 机器学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]