Characteristics of color discrimination on a face image

Journal of Vision(2019)

引用 0|浏览10
暂无评分
摘要
Skin color is essential for obtaining various information on our mind and body such as health, age, and face impression. Previous studies have reported that people show high sensitivity to changes in the redness of the skin. It was also reported that a reddish skin looks whiter than yellowish skin (Yoshikawa et al. 2012). These results suggest that people have a unique perception of the face and skin color. This perception specific to face and skin color would be due to the property of skin color determined by pigments components. In this study, we examined the color discrimination of face color in the direction of change in the amount of melanin and hemoglobin which are main pigment components consisting of skin color. We also tested the color discrimination on the a*b* axes in the CIELAB space. We used a Japanese female face with an averaged skin color obtained by measuring the skin of 694 Japanese females as a reference stimulus. Color changes due to melanin and hemoglobin change in the CIELAB space was calculated based on skin reflectance data obtained by the Monte Carlo Simulation (MCS) and the spectra of the D65 illuminant. Then, the color of a skin image was modulated in the direction of increase and decrease of each pigment or a*b* axes, and they used as test stimuli. A reference stimulus and a test stimulus were presented side by side on a CRT monitor. Observers adjusted the color of the test stimulus along one of modulation directions and determined a boundary which was discriminable from the reference stimulus. Our result showed better discrimination for changes in reddish direction in both the pigment change and a*b* axes, suggesting that people have more sensitive to color change accompanying the increase of hemoglobin.
更多
查看译文
关键词
color discrimination,face,image,characteristics
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要