基于图像的舌体胖瘦分析

Analysis based on the image of the tongue body fat and thin

舌诊是我国传统医学四诊"望、闻、问、切"中的重要内容,但是传统的舌诊方法主要依靠医生目视,主观观察诊断病情。这种方法缺乏客观评价依据,制约了舌诊的进一步应用和发展。文章首先利用计算机技术对舌图像实现自动分割,然后对分割后的舌体使用不同的曲线拟合函数,分析舌体胖瘦,并比较不同拟合函数之间的准确率。此算法对舌体胖瘦的识别有较高的准确率,高斯曲线拟合准确率高于二次函数曲线拟合。

Objective Tongue inspection is the main content of four diagnostic methods ＂obselation,auscultation and olfaction, inlerrngalion,palpation for traditinnal Chinese medical science. Traditional tongue inspection mainly depends on doetorg eyesight and subjective observe to diagqlose a disease. It lacks objectivity and hinders further development. In this paper, using computer lechnolngy for automatic tongue image segmentation,and then split the tongue using a variety of curve fitting function,analysis of tongue weight,accuracy and comparison of different. This algorithm to identify the body fat, thin tongue has high accuracy rate, Gaussian curve fitting is more accuracy than quadratic curve fitting.