自然影像中的臉部區域通常是視覺評估的主要物件，膚色分析對於提升影像品質扮演重要的一部分。再現喜好膚色對於色彩應用系統的色彩再現是重要的因素之一，目前有許多膚色再現的方法已被提出。
在色彩再現中，膚色調整可以根據主觀的評估方法來設計，因為觀測者的喜好對於影像品質評估是一重要的量測，這暗示了觀測者對特定影像的喜好滿意程度。因此，根據主觀評估結果，可將膚色透過色彩轉換至喜好的膚色，使得影像品質可有效被提升。所以找出喜好的膚色對於色彩再現是很重要的部分。
本研究設計了三個心理物理實驗，透過東方實驗受測者來決定東方女性之喜好膚色在sRGB規格顯示器上的色度值和亮度值範圍。第一個實驗主要是探討測試影像在顯示器上不同的色溫與亮度設定對喜好膚色的影響。其結果表示受測者對膚色的喜好程度並不受影像內容與顯示器之色溫與亮度的不同設定而改變。
第二個實驗則是透過實驗一的結果找出更精確的喜好膚色。並透過喜好分數的分析，建立且在CIELAB a*b*平面上畫出喜好分佈圖。而最後一個實驗則是探討在顯示器上人臉膚色的明度範圍與感知喜好程度的關係，在此透過臉部膚色之明度成分變化探討對喜好程度的影響。最後，本研究提出一色彩選擇的方法以製作影像從屬調色盤。此調色盤設計可方便地透過色彩量測或視覺評估來分析顯示影像的喜好色，同時亦可再現喜好色。
本研究找出東方女性之最喜好膚色色度值 (C*ab, hab) 為 (25.4, 50.5) 與明度值L*為65。觀察者對於膚色偏好的差異上，色度成分比色相成分差異大，平均∆E*ab 單位差異為3.7。另外亦發現，在本研究所探討之接近日光的白點範圍與亮度範圍上，CIELAB在色適應上表現良好。本研究並提供了一個有效檢測影像品質與再現喜好膚色的方法。

For applying natural images, the facial area is often the key object to be analyzed. Reproducing pleasing skin colors is one of important factors for achieving high image quality of color imaging systems. Hence, many methods to produce pleasing skin color have already been proposed.
For color reproduction, preferred adjustment of skin color can be established according to subjective evaluation. Observer preference is an important measure for image quality evaluation, and indicates the degree of observer satisfaction with a specific image. Subsequently, the image quality can be improved effectively by means of transforming the skin colors in an image to the preferred skin colors. It is important to realize preferred skin color for preference color reproduction.
Three psychophysical experiments were conducted to determine the preferred skin color ranges of East Asian women by East Asian observers on the chromatic and lightness components of CIELAB color space displayed on a sRGB-gamut display. In the first experiment, the preferred skin color was sought with different settings including correlated color temperatures (CCTs) and luminance levels of the display. The results indicate that the most preferred skin color on a display had small differences between various images varying in CCT, and luminance level.
In the second experiment, a more accurate center of preferred skin color was produced. The preference scores were analyzed and the preference map of skin color in CIELAB a*b* plane was established. The aim of the last experiment was to examine the relationship between the lightness range of facial skin color and the perceived preference on a display. It was conducted to examine the effect of the lightness component on the preference of facial skin color. Finally, image-dependent color palette is proposed to apply to preferred skin color reproduction. It will be useful to analyze skin colors of displayed images, by friendly performances of color measurement, visual assessment and preferred color reproduction at a time.
The current study reveals that the (C*ab, hab) value of the most preferred skin colors of East Asian women is (25.4, 50.5) with an L* value of 65. The observer variances in skin color preference are larger in chroma component than in hue component with a mean ∆E*ab units of 3.7. It was also found that CIELAB performs well as a chromatic adaptation transform for all the whites close to the white points set close to daylight illuminants with varying luminance levels studied. And an effective solution was provided for testing image quality of color display systems and reproducing preferred skin colors.

[1] Lee E.J. and Ha Y.H. (1998). Favorite Color Correction for Reference Color. IEEE Transactions on Consumer Electronics, 44, 10-15.
[2] Zeng H. (2011). Preferred Skin Colour Reproduction. Unpublished doctoral dissertation, University of Leeds, United Kingdom.
[3] Kim D.H., Do H.C. and Chien S.I. (2005). Preferred Skin Color Reproduction Based on Adaptive Affine Transform. IEEE Transactions on Consumer Electronics, 51, 191-197.
[4] Graham Saxby (2010). The science of imaging: An introduction (2nd ed.). United Kingdom: Taylor & Francis.
[5] Richard D. Zakia (2002). Perception and imaging (2nd ed.). United Kingdom: Focal Press.
[6] Lee H.C. (2005). Introduction to color imaging science. New York, United States of America: Cambridge University Press.
[7] Ohta N. and Robertson A. (2005). Colorimetry: Fundamentals and applications. United Kingdom: John Wiley & Sons Ltd.
[8] Stephen Westland and Caterina Ripamonti (2004). Computational colour science using MATLAB. United Kingdom: John Wiley & Sons Ltd.
[9] Sueeprasan S. (2003). Evaluation of colour appearance models and indices for evaluation of daylight simulators to provide predictable cross-media colour reproduction. Unpublished doctoral dissertation, University of Derby, United Kingdom.
[10] Li C., Luo M.R. and Hunt R.W.G. (2000). A Revision of the CIECAM97s Model. Color Research and Application, 25, 260-266.
[11] Li C., Luo M.R., Rigg B. and Hunt R.W.G. (2002). CMC 2000 Chromatic Adaptation Transform: CMCCAT2000. Color Research and Application, 27, 49-58.
[12] Luo M.R., Rigg B. and Smith K. J. (2003). CMC 2002 Colour Inconstancy Index; CMCCON02. Color Technology, 119, 280–285.
[13] Ján Morovič (2008). Color gamut mapping. United Kingdom: John Wiley & Sons Ltd.
[14] Hunt R. W. G., Pitt I.T. and Winter L.M. (1974). The Preferred Reproduction of Blue Sky, Green Grass and Caucasian Skin in Colour Photography. Journal of Photograhphic Science, 22, 144-149.
[15] Hunt R. W. G. (2004). Reproduction of Colour (6th Ed.). United Kingdom: John Wiley & Sons Ltd.
[16] Sanger D., Asada T., Haneishi H. and Miyake Y. (1994). Facial Pattern Detection and its Preferred Color Reproduction. IS&T/SID's 2nd Color Imaging Conference, 149-153.
[17] Bartleson C.J. (1959). Some Observations on the Reproduction of Flesh Colors. Photographic Science and Engineering, 3, 114-117.
[18] Bartleson C. J. (1960). Memory Colors of Familiar Objects. Journal of the Optical Society of America, 50, 73-77.
[19] Bartleson C.J. and Bray C.P. (1962). On the Preferred Reproduction of Flesh, Blue-sky, and Green-grass Colors. Photographic Science and Engineering, 6, 19–25.
[20] Novick S. B. (1972). Preferred Caucasian Skin Colour in Colour Television. British Kinematography Sound and Television, 54, 130-135.
[21] Yano T. and Hashimoto K. (1997). Preference for Japanese Complexion Color under Illumination. Color Research and Application, 22, 269-274.
[22] Kuang J., Jiang X., Quan S. and Chiu A. (2005). A psychophysical study on the influence factors of color preference in photographic color reproduction. Proc. SPIE/IS&T Electronic Imaging: Image Quality and System Performance II, 5668, 12-19.
[23] Zeng H. and Luo M.R. (2013). Colour and Tolerance of Preferred Skin Colours on Digital Photographic Images. Color Research and Application, 38, 30-45.
[24] Zeng H., and Luo M.R. (2011). Preferred Skin Colours of Africans, Caucasians, and Orientals. Proc IS&T/SID 19th Color Imaging Conference, 211-216.
[25] Yoshikawa H., Kikuchi K., Yaguchi H., Mizokami Y. and Takata S. (2012). Effect of Chromatic Components on Facial Skin Whiteness. Color Research and Application, 37, 281-291.
[26] Bodrogi P. and Tarczali T. (2001). Colour memory for various sky, skin, and plant colours: Effect of the image context. Color Research and Application, 26, 278-289.
[27] Bodrogi P. and Khanh T. Q. (2012). Illumination, Color and Imaging: Evaluation and Optimization of Visual Displays. United Kingdom: John Wiley & Sons Ltd.
[28] Tarczali T., Park D., Bodrogi P. and Kim C. Y. (2006). Long-Term Memory Colors of Korean and Hungarian Observers. Color Research and Application, 31, 176-183.
[29] Fairchild M.D. (2005). Color appearance models. United Kingdom: John Wiley & Sons Ltd.
[30] ISO/TR 16066:2003 (E). Graphic technology — Standard object colour spectra database for colour reproduction evaluation (SOCS).
[31] Hsu Y.C. and Chen H.S. (2009). Image-Dependent Preferred Color Correction Technology Based on Skin-Based Color Space. Miyazaki, Japan: The 16th International Display Workshops (IDW’9).
[32] Chen H.S., Chen S.H., Chao Y.H., Luo M.R. and P.L. Sun (2012). Applying Image-based Color Palette for Achieving High Image Quality for Display [Electronic version]. Color Research and Application.
[33] M.R. Luo and R.W.G. Hunt, “Testing Colour Appearance Models Using Corresponding-Colour and Magnitude-Estimation Data Sets”, Color Res. Appl. 23, 147–153 (1998).
[34] Smet K., Ryckaert W.R., Pointer M.R., Deconinck G. and Hanselaer P. (2010). Colour Appearance Rating of Familiar Real Objects. Color Research and Application, 36, 192–200.
[35] RIT CIS, Munsell Color Science Laboratory (2010). Available at: http://www.cis.rit.edu/mcsl/online/Spectral/SpectralData/HumanFace/Spectral_Data/.
[36] Sun Q. and Fairchild M. (2002). Statistical Characterization of Face Spectral Reflectances and its Application to Human Portraiture Spectral Estimation. Journal of Imaging Science and Technology, 46, 498-506.
[37] Martinkauppi B. (2002). Face Colour under Varying Illumination—Analysis and Applications. Unpublished doctoral dissertation, University of Oulu, Finland.
[38] Mark D. Fairchild, “Color Appearance Models: CIECAM02 and Beyond”, Proc IS&T/SID 12th Color Imaging Conference (IS&T, Springfield, VA, 2004)
[39] Trussell H.J. and Vrhel M.J. (2008). Fundamentals of digital imaging. New York, United States of America: Cambridge University Press.
[40] Stokes M., Anderson M., Chandrasekar S. and Motta R. (2011). A Standard Default Color Space for the Internet–sRGB, available at: www.color.org/srgb.html.
[41] McAndrew A. (2004). Introduction to digital image processing with Matlab. Boston, United States of America: Thomson.
[42] Chen S.H., Chen H.S., Chao Y.H. and Luo M.R. (2010). Skin-color Palette Applying to Visual Assessment and Color Measurement of FPD. Fukuoka, Japan: The 17th International Display Workshops (IDW’10).
[43] Thomas J.B., Hardeberg J.Y., Foucherot I. and Gouton P. (2008). The PLVC Display Color Characterization Model Revisited. Color Research and Application, 33, 449-460.
[44] Howard A. and Chris R. (2000). Elementary Linear Algebra: Application Version (8th ed.). United Kingdom: John Wiley & Sons, Inc.
[45] Chen H.S., Wang T.M., Chen S.H. and Liu J.S. (2011). Skin-color Correction Method Based on Hue Template Mapping for Wide Color Gamut LCD. Color Research and Application, 36, 335-348.