A new tongue colorchecker design by space representation for precise correction

Area of Science:

• Biomedical Engineering
• Color Science
• Computer Vision

Background:

• Accurate tongue color analysis is crucial for medical diagnosis.
• Existing Munsell colorcheckers lack specificity for tongue colors, limiting correction accuracy.
• A dedicated tongue color space is needed for improved color correction.

Purpose of the Study:

• To design a novel colorchecker optimized for tongue color correction.
• To establish a tongue color space for classifying tongue and non-tongue colors.
• To enhance the accuracy of tongue color analysis in medical and research applications.

Main Methods:

• Development of a tongue color space based on a large tongue color database.
• Determination of the optimal number of colors for the colorchecker (20-24 colors).
• Proposal of greedy and clustering-based color selection methods to optimize colorchecker design.

Main Results:

• A tongue color space was established, classifying colors as tongue or non-tongue.
• The optimal number of colors for the colorchecker was determined to be 20, with 24 ultimately included.
• The clustering-based selection method outperformed the greedy method in generating the new colorchecker.
• The proposed tongue colorchecker improved correction accuracy by 48% compared to the Munsell colorchecker.

Conclusions:

• The developed space-based tongue colorchecker significantly enhances color correction accuracy.
• The tongue color space and optimized color selection methods provide a superior tool for tongue color analysis.
• This research offers a more effective solution for accurate tongue color assessment.