Geometric and Tongue-Mouth Relation Features for Morphology Analysis of Tongue Body

详细信息    查看全文

• 关键词：Tongue morphology ; Tongue ; mouth relation feature ; SVM
• 刊名：Lecture Notes in Computer Science
• 出版年：2016
• 出版时间：2016
• 年：2016
• 卷：9917
• 期：1
• 页码：490-497
• 丛书名：Advances in Multimedia Information Processing - PCM 2016
• ISBN：978-3-319-48896-7
• 卷排序：9917

文摘

Traditional Chinese Medicine diagnoses a wide range of health conditions by examining morphology features of the tongue, such as fat, thin and normal. This paper presents an approach of classification for recognizing and analyzing tongue morphology based on geometric features and tongue-mouth relation feature. The geometric features are defined using various measurements of width and length of the tongue body, and ratio between them. In addition, an innovative and important feature is proposed based on the relationship between the width of the tongue body and the width of the oral cavity, named as tongue-mouth relation feature. All these features are used to train a SVM classifier. Experimental results show that the tongue-mouth relation feature is helpful to improve the recognition accuracy for tongue morphology, and the proposed method, tested on a total of 200 tongue samples, achieved an accuracy of more than 92%.