Tongue image color recognition in traditional Chinese medicine.
- Author:
Yonggang WANG
1
;
Jie YANG
;
Yue ZHOU
;
Yiqin WANG
Author Information
1. Institute of Image Processing & Pattern Recognition, Shanghai Jiaotong University, Shanghai 200030, China. yonggangwang@sjtu.edu.cn
- Publication Type:Journal Article
- MeSH:
Color;
Diagnosis, Computer-Assisted;
Humans;
Image Interpretation, Computer-Assisted;
methods;
Medicine, Chinese Traditional;
methods;
Pattern Recognition, Automated;
methods;
Tongue;
pathology;
physiopathology
- From:
Journal of Biomedical Engineering
2005;22(6):1116-1120
- CountryChina
- Language:Chinese
-
Abstract:
The problem of color recognition of substances and coats in tongue images was decomposed into two steps, i. e. region separation and color recognition. In the first step, a new method based on the JSEG and the k-NN algorithms was proposed to separate coats and substances, where a prior template and some standard color patches were designed to assist the classification. Color recognition in a fine scale was carried out by the nearest neighbour method on the basis of the first step. The experiments showed the method proposed in this paper was simple and highly robust, and can be applied to tongue characterization effectively. In addition, the two important preprocessing algorithms namely color calibration and segmentation of tongue images were improved to obtain better results than the former.