摘要
本論文研究目的是主對數位化舌診系統進行分析，本論文所進行實驗分為四個部分，分別為「舌色分群實驗」、「色彩修正」、「舌體/背景分割和舌形判別」、「舌苔與非舌苔部位分割」。實驗通過色彩度量學、心理物理實驗等方法進行。實驗結果包含了對舌頭顏色的量化定義出屬於臺北的六種舌色：淡紅色、淺紅色、紅色、深紅色、暗紅色、紫色；不同光源下(日光、水平日光、白熾燈、冷白螢光燈、客製化螢光燈)使用七種白平衡校正的演算法(例如：灰度世界結合色適應變換、簡單白平衡法、根多項式回歸法等)進行校正且進行分析和比較其中的優劣，並總結出根多項式回歸法最好和在一般場景使用時候的取捨；使用影像分割算法對舌頭背景與前景進行分離，Kappa係數達到0.95以上。並與一般的演算法比較來說有著不錯的準確率。通過受測者問卷來判斷舌形判別演算法的準確率在64%的原因與改進的建議。兩種聚群演算(K-Means、Kernel-based Fuzzy C-Means)分割結果探討。本論文研究成果可做為未來設計遠距舌診系統之參考。

關鍵字：舌診系統、舌色、舌形判別。

Abstract
The aim of this thesis is to explore the development of digital tongue diagnosis system in Traditional Chinese Medicine (TCM). The experiment in this study is divided into the following four parts: (a) classification of tongue colors, (b) color correction of captured tongue images, (c) tongue body/background segmentation and tongue shape discrimination, and (d) segmentation of tongue coating and non-tongue coating.
The quantified classification of the tongue colors was classified by means of tongue color assessment experiment of Traditional Chinese physicians successfully. Tongue colors are classified into six color terms, which are pale red, light red, red, crimson, dark red, and purple. Five doctors assessed Munsell color charts to find the corresponding color distributions of each color term in tongue color.
Also, 7 white balance algorithms were tested in the captured tongue images under the lighting conditions of 5 types of light sources: simulated daylight, horizon daylight, illuminant incandescent, cool white fluorescent, and custom fluorescent. Moreover, the image segmentation method based on semantic segmentation was applied to make the separation of the background and foreground for the tongue image database.
We also performed the tongue shape judgement for the corrected tongue image database using two kinds of clustering algorithms (K-Means, and Kernel-based Fuzzy C-Means). Then predicted data were obtained. Besides, the tongue shape classification was also performed for the same tongue image database using the method of questionnaire investigation by a group of subjects. Then visual assessment data were obtained. Therefore, the predicted data and the visual assessment data were compared to determine which clustering algorithm is better. It is strongly expected that the conclusion in this study will contribute to the development of TCM clinical diagnosis in the future.

Key words:tongue diagnosis system,tongue color,tongue discrimination.

[1].Traditional Chinese medicine - Computerized tongue image analysis system - Part 5: Method of acquisition and expression of tongue colour and tongue coating colour,ISO/TR 20498-5:2019,2019.
[2].胡國瑞 et al., "顯示色彩工程學," 新北市: 全華圖書,2013。
[3]."Artificial intelligence," [Online]. Available: https://zh.wikipedia.org/zh-my/%E4%BA%BA%E5%B7%A5%E6%99%BA%E8%83%BD. [Accessed 24 05 2020]
[4].H. Wang and B. Zhu, Basic theories of traditional Chinese medicine. Singing Dragon, 2011.
[5].G. Maciocia, Tongue diagnosis in Chinese medicine. Eastland press Seattle, WA, 1995.
[6].G. Maciocia, "Foundations of Chinese Medicine," Elsevier Churchill Livingstone, 2015.
[7].M. Giovanni, "The foundations of Chinese medicine," Elsevier Science Health Science Div,2015.
[8].J. L. Tang, B. Y. Liu, and K. W. Ma, "Traditional Chinese medicine," Lancet, vol. 372, no. 9654, pp. 1938-1940.
[9].F. Zhao-fu and G. Yi-di, Mirror of health: tongue diagnosis in Chinese medicine. People's Medical Publishing House, 2007.
[10].I. Solos and Y. Liang, "A historical evaluation of Chinese tongue diagnosis in the treatment of septicemic plague in the pre-antibiotic era, and as a new direction for revolutionary clinical research applications," Journal of integrative medicine, vol. 16, no. 3, pp. 141-146, 2018.
[11].J. Haller Jr, "The foul tongue: a 19th century index of disease," Western Journal of Medicine, vol. 137, no. 3, p. 258, 1982.
[12].B. V. Reamy, R. Derby, and C. W. Bunt, "Common tongue conditions in primary care," American family physician, vol. 81, no. 5, pp. 627-634, 2010.
[13].R. Matison, R. Mayeux, J. Rosen, and S. Fahn, "“Tip‐of‐the‐tongue” phenomenon in Parkinson disease," Neurology, vol. 32, no. 5, pp. 567-567, 1982.
[14].H. Jeghers, "Nutrition: the appearance of the tongue as an index of nutritional deficiency," New England Journal of Medicine, vol. 227, pp. 221-228, 1942.
[15].P. De Faria, P. Vargas, P. Saldiva, G. Böhm, T. Mauad, and O. De Almeida, "Tongue disease in advanced AIDS," Oral diseases, vol. 11, no. 2, pp. 72-80, 2005.
[16].S. Xie et al., Atlas of tongue diagnosis for AIDS patients. Peoples Medical Publishing House, 2006.
[17].L.-c. Lo, Y.-F. Chen, W.-J. Chen, T.-L. Cheng, and J. Y. Chiang, "The study on the agreement between automatic tongue diagnosis system and traditional chinese medicine practitioners," Evidence-Based Complementary and Alternative Medicine, vol. 2012, 2012.
[18].J. Wu, Y. Zhang, and J. Bai, "Tongue area extraction in tongue diagnosis of traditional Chinese medicine," in 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference, 2006: IEEE, pp. 4955-4957.
[19].B. Pang, D. Zhang, N. Li, and K. Wang, "Computerized tongue diagnosis based on Bayesian networks," IEEE Transactions on biomedical engineering, vol. 51, no. 10, pp. 1803-1810, 2004.
[20].B. Zhang, X. Wang, J. You, and D. Zhang, "Tongue color analysis for medical application," Evidence-Based Complementary and Alternative Medicine, vol. 2013, 2013.
[21].I. Gareus et al., "Introducing a computer-assisted, digital tongue-imaging device for standardization of traditional tongue diagnosis," Complementary Medicine Research, vol. 21, no. 3, pp. 190-196, 2014.
[22].C.-C. Chiu, "A novel approach based on computerized image analysis for traditional Chinese medical diagnosis of the tongue," Computer methods and programs in biomedicine, vol. 61, no. 2, pp. 77-89, 2000.
[23].X. Wang, B. Zhang, Z. Yang, H. Wang, and D. Zhang, "Statistical analysis of tongue images for feature extraction and diagnostics," IEEE Transactions on Image Processing, vol. 22, no. 12, pp. 5336-5347, 2013.
[24].Y.-Z. Li et al., "Syndrome differentiation analysis on Mars500 data of traditional Chinese medicine," The Scientific World Journal, vol. 2015, 2015.
[25].L. Jiang, W. Xu, and J. Chen, "Digital imaging system for physiological analysis by tongue colour inspection," in 2008 3rd IEEE Conference on Industrial Electronics and Applications, 2008: IEEE, pp. 1833-1836.
[26].T. Oji et al., "Study of factors involved in tongue color diagnosis by Kampo medical practitioners using the Farnsworth-Munsell 100 hue test and tongue color images," Evidence-based Complementary and Alternative Medicine, vol. 2014, 2014.
[27].G. Hong, M. R. Luo, and P. A. Rhodes, "A study of digital camera colorimetric characterization based on polynomial modeling," Color Research & Application: Endorsed by Inter‐Society Color Council, The Colour Group (Great Britain), Canadian Society for Color, Color Science Association of Japan, Dutch Society for the Study of Color, The Swedish Colour Centre Foundation, Colour Society of Australia, Centre Français de la Couleur, vol. 26, no. 1, pp. 76-84, 2001.
[28].V. Cheung, S. Westland, D. Connah, and C. Ripamonti, "A comparative study of the characterisation of colour cameras by means of neural networks and polynomial transforms," Coloration technology, vol. 120, no. 1, pp. 19-25, 2004.
[29].H.-Z. Zhang, K.-Q. Wang, X.-S. Jin, and D. Zhang, "SVR based color calibration for tongue image," in 2005 International Conference on Machine Learning and Cybernetics, 2005, vol. 8: IEEE, pp. 5065-5070.
[30].L. Zhuo, P. Zhang, P. Qu, Y. Peng, J. Zhang, and X. Li, "A K-PLSR-based color correction method for TCM tongue images under different illumination conditions," Neurocomputing, vol. 174, pp. 815-821, 2016.
[31].D. Golob, M. Strafela, and V. Golob, "Color management in creating of fabric season collection," in 9th Congress of the International Colour Association, 2002, vol. 4421: International Society for Optics and Photonics, pp. 221-224.
[32].J. Shi and J. Malik, "Normalized cuts and image segmentation," IEEE Transactions on pattern analysis and machine intelligence, vol. 22, no. 8, pp. 888-905, 2000.
[33].Z. Cui, H. Zhang, D. Zhang, N. Li, and W. Zuo, "Fast marching over the 2D Gabor magnitude domain for tongue body segmentation," EURASIP journal on advances in signal processing, vol. 2013, no. 1, pp. 1-13, 2013.
[34].M. Sonka, V. Hlavac, and R. Boyle, Image processing, analysis, and machine vision. Cengage Learning, 2014.
[35].M.-J. Shi, G.-Z. Li, F.-F. Li, and C. Xu, "Computerized tongue image segmentation via the double geo-vector flow," Chinese medicine, vol. 9, no. 1, pp. 1-10, 2014.
[36].K. Wu and D. Zhang, "Robust tongue segmentation by fusing region-based and edge-based approaches," Expert systems with applications, vol. 42, no. 21, pp. 8027-8038, 2015.
[37].Y. Wang, Y. Zhou, J. Yang, and Y. Wang, "JSEG based color separation of tongue image in traditional Chinese medicine," in Iberoamerican Congress on Pattern Recognition, 2004: Springer, pp. 503-508.
[38].B. Pang, D. Zhang, and K. Wang, "The bi-elliptical deformable contour and its application to automated tongue segmentation in Chinese medicine," IEEE transactions on medical imaging, vol. 24, no. 8, pp. 946-956, 2005.
[39].S. Yamamoto et al., "Regional image analysis of the tongue color spectrum," International Journal of Computer Assisted Radiology and Surgery, vol. 6, no. 1, pp. 143-152, 2011.
[40].Y. Wang, Z. Yue, Y. Jie, and Y. Wang, "JSEG Based Color Separation of Tongue Image in Traditional Chinese Medicine," in Progress in Pattern Recognition, Image Analysis and Applications, 9th Iberoamerican Congress on Pattern Recognition, CIARP 2004, Puebla, Mexico, October 26-29, 2004, Proceedings, 2004.
[41].M.-C. Hu, M.-H. Cheng, and K.-C. Lan, "Color correction parameter estimation on the smartphone and its application to automatic tongue diagnosis," Journal of medical systems, vol. 40, no. 1, p. 18, 2016.
N. Otsu, "A threshold selection method from gray-level histograms," IEEE transactions on systems, man, and cybernetics, vol. 9, no. 1, pp. 62-66, 1979.
[42].T. Kawanabe et al., "Quantification of tongue colour using machine learning in Kampo medicine," European Journal of Integrative Medicine, vol. 8, no. 6, pp. 932-941, 2016.
[43]."Official Site of Munsell Color," [Online], Available:https://munsell.com/about-munsell-color/why-munsell/,[Accessed 23 05 2020]
V. J. Easton and J. H. McColl, "Statistics Glossary v1. 1," 1997.
[44]."Richard Brown ,"fitellipse.m"," [Online], Available:https://www.mathworks.com/matlabcentral/fileexchange/15125-fitellipse-m,[Accessed 17 06 2020]
M. Ebner, "Evolving color constancy," Pattern Recognition Letters, vol. 27, no. 11, pp. 1220-1229, 2006.
[45]. J. Von Kries, "Chromatic adaptation [excerpts](from Festschrift der Albrecht-Ludwigs-Universitat 1902; reprinted from Sources of Color Science, pp. 109-119)," SPIE MILESTONE SERIES MS, vol. 77, pp. 47-47, 1993.
[46].K. Lam, "Metamerism and colour constancy, 1995," PhD thesis.
[47].G. D. Finlayson and S. Süsstrunk, "Spectral sharpening and the Bradford transform," in Proc. Color Imaging Symposium (CIS 2000), 2000, no. CONF, pp. 236-243.
[48].C. Li, M. R. Luo, B. Rigg, and R. W. Hunt, "CMC 2000 chromatic adaptation transform: CMCCAT2000," Color Research & Application: Endorsed by Inter‐Society Color Council, The Colour Group (Great Britain), Canadian Society for Color, Color Science Association of Japan, Dutch Society for the Study of Color, The Swedish Colour Centre Foundation, Colour Society of Australia, Centre Français de la Couleur, vol. 27, no. 1, pp. 49-58, 2002.
[49].N. Moroney, M. D. Fairchild, R. W. Hunt, C. Li, M. R. Luo, and T. Newman, "The CIECAM02 color appearance model," in Color and Imaging Conference, 2002, vol. 2002, no. 1: Society for Imaging Science and Technology, pp. 23-27.
[50].N. Limare, J.-L. Lisani, J.-M. Morel, A. B. Petro, and C. Sbert, "Simplest color balance," Image Processing On Line, vol. 1, pp. 297-315, 2011.
[51].G. D. Finlayson, M. Mackiewicz, and A. Hurlbert, "Color correction using root-polynomial regression," IEEE Transactions on Image Processing, vol. 24, no. 5, pp. 1460-1470, 2015.
[52].L.-C. Chen, Y. Zhu, G. Papandreou, F. Schroff, and H. Adam, "Encoder-decoder with atrous separable convolution for semantic image segmentation," in Proceedings of the European conference on computer vision (ECCV), 2018, pp. 801-818.
[53].T.-Y. Lin et al., "Microsoft coco: Common objects in context," in European conference on computer vision, 2014: Springer, pp. 740-755.
[54].B. Zhang and H. Zhang, "Significant geometry features in tongue image analysis," Evidence-based Complementary and Alternative Medicine: eCAM, vol. 2015, 2015.
[55].J. MacQueen, "Some methods for classification and analysis of multivariate observations," in Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, 1967, vol. 1, no. 14: Oakland, CA, USA, pp. 281-297.
[56].J. C. Bezdek, Pattern recognition with fuzzy objective function algorithms. Springer Science & Business Media, 2013.
[57].D. Zhang, K. Tan, and S. Chen, Semi-supervised Kernel-Based Fuzzy C-Means. Springer Berlin Heidelberg, 2004.
[58].Traditional Chinese medicine—Computerized tongue image analysis system—Part1:General requirements,ISO,20498-1:2018,2019.
[59]."X-RITE (MACBETH) SPECTRALIGHT III VIEWING BOOTH," [Online]. Available:https://www.rpimaging.com/x-rite-spectralight-iii-viewing-booth.html,[Accessed 24 05 2020].
[60]."ColorChecker Classic Mini," [Online]. Available:https://www.xrite.com/CATegories/calibration-profiling/colorchecker-targets/colorchecker-classic-mini,[Accessed 24 05 2020].
[61]."ColorEdge CG246," [Online]. Available:https://www.eizo.com/products/coloredge/cg246/,[Accessed 24 05 2020].
[62]."i1Basic Pro 2," [Online]. Available:https://www.xrite.com/CATegories/calibration-profiling/i1basic-pro-2,[Accessed 24 05 2020].
[63].李乃民et al., "舌診學," 學苑出版社, 2006。
[64].陳魯南, "織色入史," 中華書局,2014。
T. Oji, T. Namiki, K. Mitani, K. Ueda, and T. Hanawa, "Japanese Literature Survey of Tongue Findings for the Purpose of Creating a Unified Multicenter Description of Clinical Tongue Diagnoses," Kampo Medicine, vol. 65, no. 3, pp. 224-230, 2014.
[66].千葉大學醫學研究所和漢診療學講座, "舌診臨床診斷記載," 日本千葉大學醫學部,2012.11.01 Ver.1。