Facial redness is an important perceptual attribute that receives many concerns from application fields such as dermatology and cosmetics. Existing studies have commonly used the average CIELAB a* value of the facial skin area to represent the overall facial redness. Yet, the perception of facial redness has never been precisely examined. This research was designed to quantify the perception of facial redness and meanwhile investigate the perceptual difference between the faces and the uniform patches. Eighty images of real human faces and uniform skin colour patches were scaled in terms of their perceived redness by a panel of observers. The results showed that the CIELAB a* was not a good predictor of facial redness since the perceived redness was also affected by the L* and b* values. A new index, RIS was developed to accurately quantify the perception of facial skin redness, which promised a much higher accuracy (R2 = 0.874) than the a* value (R2 = 0.461). The perceptual differenc between facial redness and patch redness was also discussed.
Yan Lu, Kaida Xiao, Cheng Li, "Facial Redness Perception based on Realistic Skin Models" in London Imaging Meeting, 2023, pp 76 - 80, https://doi.org/10.2352/lim.2023.4.1.19