Using Color Morphological Parameters for Recognition of Tongue Coating Properties

• Main Authors: Chia-chi Wang, 王嘉麒
• Other Authors: Chuang-chien Chiu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/50109014624572991797

碩士 === 逢甲大學 === 自動控制工程所 === 95 === The traditional Chinese medicine has been convinced to be promoted by modern approaches in more than three decades. For instance, using a computerized tongue examination system based on digital image processing techniques for the purpose of quantifying the tongue properties and building the database for patients are two important and desirable trends. In this research, one purpose is to modify the computerized tongue examination system, which provides a well calibrated tongue images capture environment. In previous research, the quantitative analysis of tongue coating properties applied both the gray level tongue image and color tongue image to get the texture values of gray level image array and the spectrum of Fourier transform. However, the relative coordinates of image could not be depicted and some useful texture information might lose. The other purpose of this research is to retain the color information in tongue for quantitative analysis of tongue coating by using morphology. We will add and substrate the color values of each pixel to connect with the coordinate axis. The results has demonstrated to be more helpful for the quantitative analysis of tongue coating. This recognition rate has reached 87.5% in classifying the thick coatings and the thin coatings. But the recognition of grimy coatings and crusted coatings can only interpret the variation in tongue coating because the tongue images are not enough to make a conclusion. We expect to assist the clinicians in the evaluation of medical treatment and it will also enhance the quality of medical treatment in the future.