Abstract
A total absence of gloss can make the skin appear dry, while excessive gloss can contribute to an unattractive, “oily-look.” Traditionally, skin gloss has been measured by the intensity of surface reflection. The disadvantage of this method is that the gloss is determined by the amount of the surface reflection in small, localized facial areas only, and does not permit evaluation of the overall optical impression of the entire face. In our work, we recognized differences in the optical contribution of facial features such as the nose and cheekbones, and surface characteristics such as pores and wrinkles. Highlights from facial features are usually regarded as normal whereas the highlight from skin texture can accentuate the presence of large pores and wrinkles, contributing to a negative impression. For an accurate evaluation of facial gloss impression, these contributions were taken into consideration in designing our system. In this study, we propose a new method of evaluating facial gloss using an image processing technique. By this method, the surface reflection image is computed from the digital color image, and separated into the facial feature and texture images for quantitative evaluation. The purpose of this work is to explore the possibility of using 2-dimensional parameters of gross appearance in our system for more accurate and objective evaluation of gloss impression.
