Digital tongue color imaging and analyzing for Traditional Chinese medicine applications

Pao Yue-kong Library Electronic Theses Database

Digital tongue color imaging and analyzing for Traditional Chinese medicine applications


Author: Wang, Xingzheng
Title: Digital tongue color imaging and analyzing for Traditional Chinese medicine applications
Degree: Ph.D.
Year: 2013
Subject: Tongue manifestations of general diseases -- Data processing.
Medicine, Chinese -- Data processing.
Hong Kong Polytechnic University -- Dissertations
Department: Dept. of Computing
Pages: xviii, 159 p. : ill. (chiefly col.) ; 30 cm.
Language: English
InnoPac Record:
Abstract: With the development of digital image processing and pattern classification technology, computer-aided tongue image analysis for medical application has been found to be an effective way to solve the intrinsic drawbacks lie in traditional tongue diagnosis which is subjective and experience-based. Although great progress has been achieved in this research topic, little attention has been paid on the color imaging, color correction and analyzing technologies. In this regard, this thesis explored different aspects of digital color imaging and analysis techniques. First, a novel tongue imaging system which can record tongue information faithfully and precisely for medical analysis was developed. During the implementation of this device, a thorough requirement analysis was conducted firstly to summarize the demanding of rendering tongue image reliably, and a series of criteria were presented to fulfill these requirements thereby to accurately and consistently extracted all possible image features. Afterwards, modules of the device were developed accordingly by following these guidelines. Performance tests showed that this device could acquire images in high quality and consistency under various environmental conditions. Second, an optimized correction scheme is proposed to correct variations of images caused by system components thus to achieve accurate tongue image rendering. The correction algorithm in this scheme is generated by comparing several popular correction algorithms, i.e., polynomial-based regression, ridge regression, support vector regression, and neural network mapping algorithms. Experimental results showed that the CIELAB color difference between the *corrected value and standard values is less than 5 (ΔE*ab < 5) which is hardly distinguished by the human eye.
Then, an in-depth analysis on the space distribution characteristics of tongue colors which aims to propose a statistically described tongue color space was presented. Three aspects of this color space, i.e., tongue color gamut which defines the range of human tongue colors, color centers of 12 tongue color categories, and color distribution of typical image features in the tongue color gamut, were elaborately investigated in this study. This proposed tongue color space is the most comprehensive and detailed study on tongue color space to date, and it is essential and critical for objective tongue color feature extraction for medical application. After that, based on the above tongue color space, a novel tongue color feature extraction method was proposed, and tongue images were classified into different health condition groups. The general principle for this color feature extraction is to compare pixels’ color value of the input tongue image with the 12 colors centers, and assign the nearest color category to these pixels. Experimental results on a large tongue image database showed that a given tongue sample can be classified into health and disease with an average accuracy of 82.35%. Further testing showed image samples from over ten types of disease can be classified at an average accuracy of 70%. Finally, in order to further improve the correction accuracy, a novel colorchecker which is dedicated for tongue color correction was developed. Three essential issues leading to the development of this colorchecker, i.e., where to choose checker colors, how many colors should be involved and how to select the best combination of color samples, were elaborately investigated in this study. Compared to Munsell colorchecker, this proposed space-based colorchecker can greatly improve the correction accuracy by 48%. Further experimental results on more correction tasks also validated its effectiveness and superiority.

Files in this item

Files Size Format
b26160262.pdf 6.329Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.


Quick Search


More Information